[Ebooks PDF] download Machine Learning for Practical Decision Making: A Multidisciplinary Perspective with Applications from Healthcare, Engineering and Business Analytics Christo El Morr full chapters

Full download ebooks at https://ebookmeta.
com
Machine Learning for Practical Decision Making:

A Multidisciplinary Perspective with
Applications from Healthcare, Engineering and
Business Analytics Christo El Morr
For dowload this book click link below
https://ebookmeta.com/product/machine-learning-for-
practical-decision-making-a-multidisciplinary-perspective-
with-applications-from-healthcare-engineering-and-business-
analytics-christo-el-morr/
OR CLICK BUTTON
DOWLOAD NOW
More products digital (pdf, epub, mobi) instant
download maybe you interests ...
Machine Learning for Business Analytics: Concepts,

Techniques and Applications with JMP Pro, 2nd Edition
Galit Shmueli
https://ebookmeta.com/product/machine-learning-for-business-
analytics-concepts-techniques-and-applications-with-jmp-pro-2nd-
edition-galit-shmueli/
Machine Learning for Business Analytics: Concepts,

Techniques and Applications with JMP Pro, 2nd Edition
Galit Shmueli
https://ebookmeta.com/product/machine-learning-for-business-
analytics-concepts-techniques-and-applications-with-jmp-pro-2nd-
edition-galit-shmueli-2/
Business Analytics A Data Driven Decision Making

Approach for Business Volume I 1st Edition Amar Sahay
https://ebookmeta.com/product/business-analytics-a-data-driven-
decision-making-approach-for-business-volume-i-1st-edition-amar-
sahay/
Deep Learning Applications and Intelligent Decision

Making in Engineering 1st Edition Balamurugan Shanmugam
(Editor)
https://ebookmeta.com/product/deep-learning-applications-and-
intelligent-decision-making-in-engineering-1st-edition-
balamurugan-shanmugam-editor/
Machine Learning for Decision Makers: Cognitive
Computing Fundamentals for Better Decision Making 2nd
Edition Patanjali Kashyap
https://ebookmeta.com/product/machine-learning-for-decision-
makers-cognitive-computing-fundamentals-for-better-decision-
making-2nd-edition-patanjali-kashyap/
Practical AI for Healthcare Professionals: Machine

Learning with Numpy, Scikit-learn, and TensorFlow 1st
Edition Abhinav Suri
https://ebookmeta.com/product/practical-ai-for-healthcare-
professionals-machine-learning-with-numpy-scikit-learn-and-
tensorflow-1st-edition-abhinav-suri/
AI and Society Tensions and Opportunities Chapman Hall

CRC Artificial Intelligence and Robotics Series 1st
Edition Christo El Morr
https://ebookmeta.com/product/ai-and-society-tensions-and-
opportunities-chapman-hall-crc-artificial-intelligence-and-
robotics-series-1st-edition-christo-el-morr/
Developing Informed Intuition for Decision Making Data

Analytics Applications 1st Edition Liebowitz Jay
https://ebookmeta.com/product/developing-informed-intuition-for-
decision-making-data-analytics-applications-1st-edition-
liebowitz-jay/
Spreadsheet Modeling and Decision Analysis: A Practical

Introduction to Business Analytics, 9e Cliff Ragsdale
https://ebookmeta.com/product/spreadsheet-modeling-and-decision-
analysis-a-practical-introduction-to-business-analytics-9e-cliff-
ragsdale/
International Series in
Operations Research & Management Science
Christo El Morr
Manar Jammal
Hossam Ali-Hassan
Walid EI-Hallak
Machine Learning
for Practical
Decision Making
A Multidisciplinary Perspective
with Applications from Healthcare,
Engineering and Business Analytics
International Series in Operations Research &
Management Science
Founding Editor
Frederick S. Hillier, Stanford University, Stanford, CA, USA
Volume 334
Series Editor
Camille C. Price, Department of Computer Science, Stephen F. Austin State Uni-
versity, Nacogdoches, TX, USA
Editorial Board Members

Emanuele Borgonovo, Department of Decision Sciences, Bocconi University,
Milan, Italy
Barry L. Nelson, Department of Industrial Engineering & Management Sciences,
Northwestern University, Evanston, IL, USA
Bruce W. Patty, Veritec Solutions, Mill Valley, CA, USA
Michael Pinedo, Stern School of Business, New York University, New York, NY, USA
Robert J. Vanderbei, Princeton University, Princeton, NJ, USA
Associate Editor
Joe Zhu, Foisie Business School, Worcester Polytechnic Institute, Worcester, MA, USA
The book series International Series in Operations Research and Management
Science encompasses the various areas of operations research and management
science. Both theoretical and applied books are included. It describes current
advances anywhere in the world that are at the cutting edge of the field. The series
is aimed especially at researchers, advanced graduate students, and sophisticated
practitioners.
The series features three types of books:
• Advanced expository books that extend and unify our understanding of partic-
ular areas.
• Research monographs that make substantial contributions to knowledge.
• Handbooks that define the new state of the art in particular areas. Each
handbook will be edited by a leading authority in the area who will organize a
team of experts on various aspects of the topic to write individual chapters. A
handbook may emphasize expository surveys or completely new advances (either
research or applications) or a combination of both.
The series emphasizes the following four areas:
Mathematical Programming: Including linear programming, integer program-
ming, nonlinear programming, interior point methods, game theory, network opti-
mization models, combinatorics, equilibrium programming, complementarity
theory, multiobjective optimization, dynamic programming, stochastic program-
ming, complexity theory, etc.
Applied Probability: Including queuing theory, simulation, renewal theory,
Brownian motion and diffusion processes, decision analysis, Markov decision
processes, reliability theory, forecasting, other stochastic processes motivated by
applications, etc.
Production and Operations Management: Including inventory theory, produc-
tion scheduling, capacity planning, facility location, supply chain management,
distribution systems, materials requirements planning, just-in-time systems, flexible
manufacturing systems, design of production lines, logistical planning, strategic
issues, etc.
Applications of Operations Research and Management Science: Including
telecommunications, health care, capital budgeting and finance, economics, market-
ing, public policy, military operations research, humanitarian relief and disaster
mitigation, service operations, transportation systems, etc.
This book series is indexed in Scopus.
Christo El Morr • Manar Jammal •
Hossam Ali-Hassan • Walid El-Hallak
Machine Learning
for Practical Decision Making
A Multidisciplinary Perspective with
Applications from Healthcare, Engineering
and Business Analytics
Christo El Morr Manar Jammal
School of Health Policy and Management School of Information Technology
York University York University
Toronto, ON, Canada Toronto, ON, Canada
Hossam Ali-Hassan Walid El-Hallak

Department of International Studies Ontario Health
York University, Glendon Campus Toronto, ON, Canada
Toronto, ON, Canada
ISSN 0884-8289 ISSN 2214-7934 (electronic)

International Series in Operations Research & Management Science
ISBN 978-3-031-16989-2 ISBN 978-3-031-16990-8 (eBook)
https://doi.org/10.1007/978-3-031-16990-8
© The Editor(s) (if applicable) and The Author(s), under exclusive license to Springer Nature Switzerland
AG 2022
This work is subject to copyright. All rights are solely and exclusively licensed by the Publisher, whether
the whole or part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way, and
transmission or information storage and retrieval, electronic adaptation, computer software, or by
similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this publication
does not imply, even in the absence of a specific statement, that such names are exempt from the relevant
protective laws and regulations and therefore free for general use.
The publisher, the authors, and the editors are safe to assume that the advice and information in this
book are believed to be true and accurate at the date of publication. Neither the publisher nor the authors or
the editors give a warranty, expressed or implied, with respect to the material contained herein or for any
errors or omissions that may have been made. The publisher remains neutral with regard to jurisdictional
claims in published maps and institutional affiliations.
This Springer imprint is published by the registered company Springer Nature Switzerland AG
The registered company address is: Gewerbestrasse 11, 6330 Cham, Switzerland
To our families for their love and support. To
our students, they are our inspiration.
Preface
This book fills a gap in the machine learning literature. It presents the machine
learning concepts in a very simple and practical way starting with a tangible example
and building on it the theory. Each machine learning chapter starts with a presenta-
tion of the problem to solve, followed by a practical example on how the algorithm
that solves the problem works, then comes the presentation of the machine learning
algorithm theory and closing remarks (e.g., pitfalls, advantages limitations). This is
followed by a set of key terms, a set of questions to test your understanding of the
chapter material, a set of references to read more about the subject and hands-on lab
exercises. The lab exercises allow readers to apply the chapter’s concepts and gain
needed skills. To maximize the benefit for readers and to expose them to a myriad of
machine learning languages and frameworks, the lab exercises (and sometimes the
chapter’s material) are built around Python mainly, followed by R (and R Studio),
and Weka. One of the strengths of this book is that it can be used by people who are
exposed to programming or would like to learn how to program and those who prefer
not to program but to solve decision-making problems with machine learning using
simple graphical user interface. The former can use Python, the machine learning
language par excellence, throughout all chapters, or R (limited to Chaps. 1–4, and 6).
The use of Tableau, a visual analytics platform, is reserved to Chap. 5, Data
Visualization, while Weka is explained and used in Chaps. 4 and 6–12; given the
simplicity of Weka, we believe that there was no necessity to add Weka-based lab
exercises for Chaps. 13–15.
This introductory textbook to machine learning for decision making can be used
by students in Computer Science, Information Technology, Health Informatics, and
Business fields. Depending on the students’ level of study and exposure to technol-
ogy, either Weka or Python can be used. However, given the pervasive use of Python
in the market we advise students of all sorts to get exposure to Python and how it
works. Teaching the whole Python language is beyond the scope of this book;
however, we cover Python’s libraries related to machine leaning (e.g., Scikit-
Learn, TensorFlow, Keras) and many Python programming concepts.
vii
viii Preface
Another strength of the book is its focus on the necessary content for an
introductory course to machine learning while providing enough complexity without
being complex or introducing heavy mathematical formulation; the exception being
neural networks where we considered that providing less simple mathematical
formulations was necessary as an illustration but skipping them would not be a
problem to understand the algorithm. There is no mathematical knowledge needed to
read and use this book. Chapter 2 provides, in a simple manner coupled with many
examples, the main mathematical concepts needed to understand the chapters.
A final strength of this book is the use of a variety of datasets from several
domains (e.g., health, business, social media, census, survey) which provide a good
exposure to the myriad types of applications in which machine learning can be used.
The book is organized in three parts: Part 1 is an introduction section that
encompasses Chaps. 1–5, it introduces machine learning fundamentals and allows
for installing the different software tools and the introduction to Python, R, and
Weka. The machine learning algorithms and corresponding lab exercises are covered
in Chaps. 6–15, and future perspectives are provided in Chap. 16.
Professors who adopt this book have flexibility in the way they want to teach the
material; it all depends on the objectives of their course. Some can use to teach
machine learning using Python and hence need to cover Chaps. 1, 4 and 6–16; this
could be true for courses related to practical machine learning. Professors who are
interested in teaching Analytics and (some) machine learning can cover Chaps. 1–5
and then some of the Chaps. 6–16. It is still feasible to cover all chapters in 12 weeks,
Chaps. 1 and 2 are an introduction and a quick overview that can be covered with
their labs in one session; Chaps. 3 (or some of it) and Chap. 4 are possible to
combine in one session. Chapters 5 and 16 can be covered in one session; then each
Chaps. 6–14 in one session; and final Chaps. 15 and 16 in one session. This is not to
impose a single way of approaching the textbook but to provide examples of
alternatives and demonstrate flexibility. The book provides you with flexibility to
be adopted in several contexts. The datasets used in the lab exercise touch upon
many domains: health, information technology, business, and engineering.
We hope that this first edition of the book will enrich the readers’ knowledge and
skills and we welcome their comments and suggestions. Readers can access the
Python code for the Chap. 6 to Chap. 15 lab exercises on GitHub https://github.com/
christoelmorr/ML-4-Practical-DM.git.
Toronto, ON, Canada Christo El Morr

Manar Jammal
Hossam Ali-Hassan
Walid El-Hallak
Contents
1 Introduction to Machine Learning . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.1 Introduction to Machine Learning . . . . . . . . . . . . . . . . . . . . . 1
1.2 Origin of Machine Learning . . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.3 Growth of Machine Learning . . . . . . . . . . . . . . . . . . . . . . . . . 2
1.4 How Machine Learning Works . . . . . . . . . . . . . . . . . . . . . . . 3
1.5 Machine Learning Building Blocks . . . . . . . . . . . . . . . . . . . . 4
1.5.1 Data Management and Exploration . . . . . . . . . . . . . 4
1.5.2 The Analytics Landscape . . . . . . . . . . . . . . . . . . . . 13
1.6 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
1.7 Key Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
1.8 Test Your Understanding . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
1.9 Read More . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23
1.10 Lab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.10.1 Introduction to R . . . . . . . . . . . . . . . . . . . . . . . . . . 24
1.10.2 Introduction to RStudio . . . . . . . . . . . . . . . . . . . . . . 27
1.10.3 Introduction to Python and Jupyter Notebook IDE . . . 33
1.10.4 Do It Yourself . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41
2 Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
2.1 Overview of the Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
2.2 Definition of General Terms . . . . . . . . . . . . . . . . . . . . . . . . . . 45
2.3 Types of Variables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
2.3.1 Measures of Central Tendency . . . . . . . . . . . . . . . . . 46
2.4 Inferential Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
2.4.1 Data Distribution . . . . . . . . . . . . . . . . . . . . . . . . . . 47
2.4.2 Hypothesis Testing . . . . . . . . . . . . . . . . . . . . . . . . . 47
2.4.3 Type I and II Errors . . . . . . . . . . . . . . . . . . . . . . . . 48
2.4.4 Steps for Performing Hypothesis Testing . . . . . . . . . 48
2.4.5 Test Statistics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
ix
x Contents
2.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
2.6 Key Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54
2.8 Read More . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
2.9 Lab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 55
2.9.1 Working Example in R . . . . . . . . . . . . . . . . . . . . . . 55
2.9.2 Working Example in Python . . . . . . . . . . . . . . . . . . 57
2.9.3 Do It Yourself . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
2.9.4 Do More Yourself (Links to Available Datasets for
Use) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
3 Overview of Machine Learning Algorithms . . . . . . . . . . . . . . . . . . . 61
3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
3.2 Data Mining . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
3.3 Analytics and Machine Learning . . . . . . . . . . . . . . . . . . . . . . 63
3.3.1 Terminology Used in Machine Learning . . . . . . . . . 64
3.3.2 Machine Learning Algorithms: A Classification . . . . 65
3.4 Supervised Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 66
3.4.1 Multivariate Regression . . . . . . . . . . . . . . . . . . . . . 68
3.4.2 Decision Trees . . . . . . . . . . . . . . . . . . . . . . . . . . . . 70
3.4.3 Artificial Neural Networks . . . . . . . . . . . . . . . . . . . 72
3.4.4 Naïve Bayes Classifier . . . . . . . . . . . . . . . . . . . . . . 75
3.4.5 Random Forest . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
3.4.6 Support Vector Machines (SVM) . . . . . . . . . . . . . . . 78
3.5 Unsupervised Learning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 80
3.5.1 K-Means . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
3.5.2 K-Nearest Neighbors (KNN) . . . . . . . . . . . . . . . . . . 82
3.5.3 AdaBoost . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
3.6 Applications of Machine Learning . . . . . . . . . . . . . . . . . . . . . 85
3.6.1 Machine Learning Demand Forecasting and Supply
Chain Performance . . . . . . . . . . . . . . . . . . . . . . . . . 85
3.6.2 A Case Study on Cervical Pain Assessment with
Motion Capture . . . . . . . . . . . . . . . . . . . . . . . . . . . 87
3.6.3 Predicting Bank Insolvencies Using Machine
Learning Techniques . . . . . . . . . . . . . . . . . . . . . . . . 89
3.6.4 Deep Learning with Convolutional Neural Network
for Objective Skill Evaluation in Robot-Assisted
Surgery . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
3.7 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
3.8 Key Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 92
3.10 Read More . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
3.11 Lab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
3.11.1 Machine Learning Overview in R . . . . . . . . . . . . . . 94
Contents xi
3.11.2 Supervised Learning Overview . . . . . . . . . . . . . . . . 98

3.11.3 Unsupervised Learning Overview . . . . . . . . . . . . . . 104
3.11.4 Python Scikit-Learn Package Overview . . . . . . . . . . 106
3.11.5 Python Supervised Learning Machine (SML) . . . . . . 107
3.11.6 Unsupervised Machine Learning (UML) . . . . . . . . . 111
3.11.7 Do It Yourself . . . . . . . . . . . . . . . . . . . . . . . . . . . . 113
3.11.8 Do More Yourself . . . . . . . . . . . . . . . . . . . . . . . . . 113
References . . . . ....................................... . 113
4 Data Preprocessing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
4.1 The Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 117
4.2 Data Preprocessing Steps . . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
4.2.1 Data Collection . . . . . . . . . . . . . . . . . . . . . . . . . . . 119
4.2.2 Data Profiling, Discovery, and Access . . . . . . . . . . . 119
4.2.3 Data Cleansing and Validation . . . . . . . . . . . . . . . . 119
4.2.4 Data Structuring . . . . . . . . . . . . . . . . . . . . . . . . . . . 120
4.2.5 Feature Selection . . . . . . . . . . . . . . . . . . . . . . . . . . 120
4.2.6 Data Transformation and Enrichment . . . . . . . . . . . . 120
4.2.7 Data Validation, Storage, and Publishing . . . . . . . . . 121
4.3 Feature Engineering . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 121
4.3.1 Feature Creation . . . . . . . . . . . . . . . . . . . . . . . . . . . 122
4.3.2 Transformation . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
4.3.3 Feature Extraction . . . . . . . . . . . . . . . . . . . . . . . . . 123
4.4 Feature Engineering Techniques . . . . . . . . . . . . . . . . . . . . . . . 123
4.4.1 Imputation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 123
4.4.2 Discretizing Numerical Features . . . . . . . . . . . . . . . 124
4.4.3 Converting Categorical Discrete Features to Numeric
(Binarization) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
4.4.4 Log Transformation . . . . . . . . . . . . . . . . . . . . . . . . 125
4.4.5 One-Hot Encoding . . . . . . . . . . . . . . . . . . . . . . . . . 125
4.4.6 Scaling . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 126
4.4.7 Reduce the Features Dimensionality . . . . . . . . . . . . . 128
4.5 Overfitting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 128
4.6 Underfitting . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 130
4.7 Model Selection: Selecting the Best Performing Model of an
Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 131
4.7.1 Model Selection Using the Holdout Method . . . . . . . 133
4.7.2 Model Selection Using Cross-Validation . . . . . . . . . 134
4.7.3 Evaluating Model Performance in Python . . . . . . . . . 135
4.8 Data Quality . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 136
4.9 Key Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
4.11 Read More . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 137
4.12 Lab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 138
xii Contents
4.12.2 Working Example in Weka . . . . . . . . . . . . . . . . . . . 150

4.12.3 Do It Yourself . . . . . . . . . . . . . . . . . . . . . . . . . . . . 157
4.12.4 Do More Yourself . . . . . . . . . . . . . . . . . . . . . . . . . 160
References . . . . ....................................... . 161
5 Data Visualization . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 165
5.2 Presentation and Visualization of Information . . . . . . . . . . . . . 168
5.2.1 A Taxonomy of Graphs . . . . . . . . . . . . . . . . . . . . . 168
5.2.2 Relationships and Graphs . . . . . . . . . . . . . . . . . . . . 169
5.2.3 Dashboards . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
5.2.4 Infographics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 180
5.3 Building Effective Visualizations . . . . . . . . . . . . . . . . . . . . . . 184
5.4 Data Visualization Software . . . . . . . . . . . . . . . . . . . . . . . . . . 185
5.5 Conclusion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
5.6 Key Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 187
5.8 Read More . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 188
5.9 Lab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 189
5.9.1 Working Example in Tableau . . . . . . . . . . . . . . . . . 189
5.9.2 Do It Yourself . . . . . . . . . . . . . . . . . . . . . . . . . . . . 190
5.9.3 Do More Yourself . . . . . . . . . . . . . . . . . . . . . . . . . 191
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 192
6 Linear Regression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
6.1 The Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 195
6.2 A Practical Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 196
6.3 The Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 198
6.3.1 Modeling the Linear Regression . . . . . . . . . . . . . . . 198
6.3.2 Gradient Descent . . . . . . . . . . . . . . . . . . . . . . . . . . 199
6.3.3 Gradient Descent Example . . . . . . . . . . . . . . . . . . . 202
6.3.4 Batch Versus Stochastic Gradient Descent . . . . . . . . 203
6.3.5 Examples of Error Functions . . . . . . . . . . . . . . . . . . 204
6.3.6 Gradient Descent Types . . . . . . . . . . . . . . . . . . . . . 205
6.4 Final Notes: Advantages, Disadvantages, and Best Practices . . 207
6.5 Key Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 208
6.7 Read More . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 209
6.8 Lab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 210
6.8.1 Working Example in R . . . . . . . . . . . . . . . . . . . . . . 211
6.8.4 Do It Yourself . . . . . . . . . . . . . . . . . . . . . . . . . . . . 223
6.8.5 Do More Yourself . . . . . . . . . . . . . . . . . . . . . . . . . 229
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 230
Contents xiii
7 Logistic Regression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231

7.1 The Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231
7.3 The Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 232
7.5 Key Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 235
7.7 Read More . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 236
7.8 Lab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 237
7.8.3 Do It Yourself . . . . . . . . . . . . . . . . . . . . . . . . . . . . 244
7.8.4 Do More Yourself . . . . . . . . . . . . . . . . . . . . . . . . . 247
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 248
8 Decision Trees . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
8.1 The Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 251
8.3 The Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
8.3.1 Tree Basics . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 252
8.3.2 Training Decision Trees . . . . . . . . . . . . . . . . . . . . . 259
8.3.3 A Generic Algorithm . . . . . . . . . . . . . . . . . . . . . . . 261
8.3.4 Tree Pruning . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 266
8.5 Key Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 267
8.7 Read More . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 269
8.8 Lab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 270
8.8.3 Do It Yourself . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
8.8.4 Do More Yourself . . . . . . . . . . . . . . . . . . . . . . . . . 278
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 278
9 Naïve Bayes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
9.1 The Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 279
9.2 The Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 280
9.2.1 Bayes Theorem . . . . . . . . . . . . . . . . . . . . . . . . . . . 280
9.2.2 The Naïve Bayes Classifier (NBC): Dealing with
Categorical Variables . . . . . . . . . . . . . . . . . . . . . . . 281
9.2.3 Gaussian Naïve Bayes (GNB): Dealing with
Continuous Variables . . . . . . . . . . . . . . . . . . . . . . . 283
9.3.1 Naïve Bayes Classifier with Categorical Variables
Example . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 284
xiv Contents
9.3.2 Gaussian Naïve Bayes Example . . . . . . . . . . . . . . . . 285

9.5 Key Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
9.7 Read More . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 288
9.8 Lab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 291
9.8.3 Do It Yourself . . . . . . . . . . . . . . . . . . . . . . . . . . . . 296
9.8.4 Do More Yourself . . . . . . . . . . . . . . . . . . . . . . . . . 298
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 298
10 K-Nearest Neighbors . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
10.1 The Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 301
10.2.1 A Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . 302
10.2.2 Regression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 303
10.3 The Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 305
10.3.1 Distance Function . . . . . . . . . . . . . . . . . . . . . . . . . . 305
10.3.2 KNN for Classification . . . . . . . . . . . . . . . . . . . . . . 307
10.3.3 KNN for Regression . . . . . . . . . . . . . . . . . . . . . . . . 308
10.5 Key Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 308
10.7 Read More . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 309
10.8 Lab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 311
10.8.3 Do It Yourself . . . . . . . . . . . . . . . . . . . . . . . . . . . . 317
10.8.4 Do More Yourself . . . . . . . . . . . . . . . . . . . . . . . . . 317
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 318
11 Neural Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
11.1 The Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
11.2.1 Example 1 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 319
11.3 The Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 320
11.3.1 The McCulloch–Pitts Neuron . . . . . . . . . . . . . . . . . 322
11.3.2 The Perceptron . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
11.3.3 The Perceptron as a Linear Function . . . . . . . . . . . . 326
11.3.4 Activation Functions . . . . . . . . . . . . . . . . . . . . . . . . 327
11.3.5 Training the Perceptron . . . . . . . . . . . . . . . . . . . . . . 333
11.3.6 Perceptron Limitations: XOR Modeling . . . . . . . . . . 337
11.3.7 Multilayer Perceptron (MLP) . . . . . . . . . . . . . . . . . . 337
11.3.8 MLP Algorithm Overview . . . . . . . . . . . . . . . . . . . . 340
Contents xv
11.3.9 Backpropagation . . . . . . . . . . . . . . . . . . . . . . . . . . . 342

11.3.10 Backpropagation Algorithm . . . . . . . . . . . . . . . . . . 348
11.5 Key Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 349
11.7 Read More . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 350
11.8 Lab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 352
11.8.3 Do it Yourself . . . . . . . . . . . . . . . . . . . . . . . . . . . . 357
11.8.4 Do More Yourself . . . . . . . . . . . . . . . . . . . . . . . . . 360
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 360
12 K-Means . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
12.1 The Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 361
12.3 The Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 365
12.4 Inertia . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 372
12.5 Minibatch K-Means . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374
12.7 Key Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 374
12.9 Read More . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 375
12.10 Lab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 376
12.10.2 Do It Yourself . . . . . . . . . . . . . . . . . . . . . . . . . . . . 382
12.10.3 Do More Yourself . . . . . . . . . . . . . . . . . . . . . . . . . 384
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 384
13 Support Vector Machine . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
13.1 The Problem . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 385
13.2 The Algorithm . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 386
13.2.1 Important Concepts . . . . . . . . . . . . . . . . . . . . . . . . . 388
13.2.2 Margin . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 389
13.2.3 Types of Support Vector Machines . . . . . . . . . . . . . 393
13.2.4 Classification . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 396
13.2.5 Regression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 397
13.2.6 Tuning Parameters . . . . . . . . . . . . . . . . . . . . . . . . . 397
13.2.7 Kernel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 398
13.3 Advantages, Disadvantages, and Best Practices . . . . . . . . . . . . 399
13.4 Key Terms . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 400
13.6 Read More . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 401
13.7 Lab . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 402
Another random document with
no related content on Scribd:
made. We got the facts. Our County is small, no large settlements,
and everyone knows practically everybody else, so there was little
for us to do.”
Winona, Minnesota, sends in the best kind of a report—with few or
no figures under most lettered heads. Winona has about 20,000
inhabitants, and is a small farming community with a floating
population. Much of the work of the division was in stopping local
gossip and loose talking. The League did, however, locate one
deserter, who was duly turned over.
MISSOURI
The tracing of a deserter may take a hundred pages in a file. A
certain man registered in St. Louis, but never turned in his
questionnaire. He was classified by the Adjutant General of Missouri
as a deserter, and A. P. L. was requested to find him. Search
revealed him in James City, Pennsylvania. The chief of police of a
nearby town found the man in bed. The deserter, whose name may
be called Bates, resisted fiercely. It was stated of him that he was
the first man the chief of police ever arrested who succeeded in
breaking a pair of handcuffs. He fought all the time until he was put
in jail. Mr. Bates, it is to be hoped, fought equally well in the army.
He certainly got his chance to do so.
D. W. B——, from St. Louis, was once in the 108th Infantry, but
vanished therefrom, leaving his uniform in New York with a friend.
One paragraph, the last page in the file, will cover the case of Mr. B
——: “As subject was apprehended in Buffalo, the commanding
officer at Fort Niagara was communicated with, and he detailed a
sergeant to come to Buffalo on December 17. The sergeant took B
—— into custody and conveyed him to Fort Niagara, where he is at
present.”
Kansas City, among other cases, turned in a love letter written by a
local young lady to a Japanese, Heroshirmo, at present living in
Japan. The letter begins: “Dear Heroshirmo: How I want to write to
you pages and pages of something, I am not sure what. I want to
tell you first about the beautiful summer that has just passed, how
beautiful the trees and flowers were, how infinite and blue the
sky”—but perhaps that will be enough.
The A. P. L. noticed the post-mark and thought that this sort of
correspondence ought to be looked into. It should. The Japanese
had once stopped in Kansas City as a member of a Commission on
its way to Washington, and had visited local friends. No international
plot was unveiled in this case. Just the trees and flowers were
discovered to be beautiful and the sky very blue. To be sure, the
writer being a woman, the letter had a postscript: “Just because I
have been sick, would you like to send me a genuine Japanese
kimona? I must tell you that all of the first page of your last letter
except the first few lines were cut out by the censor. D——n the
war.”
Jefferson City, Missouri, has jurisdiction over several counties but the
division consisted of only twenty-one members. These men were of
great value to the Department of Justice at Kansas City. The sparsely
settled nature of the country around Jefferson City meant a great
deal of automobile travel. The Chief says he has traveled as high as
ninety-five miles in his own car on one case. This meant a vast
amount of work for the small membership of the League at that
point. It acquitted itself admirably.
Clinton, Missouri, faithfully performed a large volume of routine work
such as comes to most of the divisions—some three hundred cases
in all, under various headings. The Chief concludes: “Our activities
have been abundant. We mean to continue our organization here
until there is no further need for it. Our personnel is made up of the
best men in this county. Our system of warning by red-white-and-
blue cards has been adopted in many States and by the National
Council of Defense.”
Monett, Missouri, had some trouble from the fact that drafted men
were at first able to obtain alcoholic beverages there. This was
stopped by the local League. There was considerable propaganda by
word of mouth in this locality which was choked off. One deserter
defied all local officers to capture him and take him back to camp.
Nevertheless he was taken, returned to camp, court-martialed and
sentenced to a term in the federal prison. As a whole, the people of
this community are law abiding 100-percent Americans. Hence the
League’s work was light.
Fayette, Missouri: “Thirty investigations resulted in reclassifying
twenty-five men. We arrested three camp deserters and two men for
disloyal acts. Found three men hoarding sugar and made them take
it back. In some cases we just warned parties that their conduct had
been reported to be reprehensible, and evidence was produced by
them to prove their later love and loyalty to the United States.”
IOWA
Des Moines, Iowa, the very prosperous capital of the prosperous
state of Iowa, had an A. P. L. man attached to the Intelligence
Service of the Army. He spoke German fluently and in order to
investigate conditions inside a neighboring camp, he pretended to be
a conscientious objector, thus being confined to barracks with other
conscientious objectors, some real and some camouflage. A picked
War Department Committee, including the Governor of the State,
was combing out these objectors and ran across the A. P. L. man.
The latter was unable to explain, and had to go through as a
conscientious objector and listen to a good lecture to boot!
Des Moines had another case of a fine looking young man who
weighed about 175 pounds and who sported a clever little military
mustache. He was caught in a slacker drive and on the following
morning hesitatingly handed the agent a telegram sent by his father,
which read: “I have told you that damned eye-brow on your upper
lip would get you into trouble. Tell the Government I say you are
only twenty—you look older, but act younger. If you wish to please
your father, enlist in the Navy.” The son enlisted.
Iowa City, Iowa, is a university town, a good, peaceful and thrifty
community and one of the most useful in the West. The foreign
element in that district has been rather Bohemian than German, but
the population has the usual admixture. There are two precincts
populated by Mennonites, whose religion is work and not war. One
of these good folk refused to buy Liberty Bonds but sold enough
walnut logs from his farm to make several thousand gun stocks. This
man was finally persuaded to buy as many dollars in bonds as his
logs made gun stocks. Some conscientious objectors from Camp
Dodge were sent out to farm among these Mennonite brothers and
thus escaped the draft, whereas local loyal farmers’ sons had to go
to the front. This created bitter feeling. Most of these dodgers were
recalled.
Oskaloosa, Iowa, had its own share of local wrangles over League
war activities. One suspect was brought up under charges of
disloyalty by reason of many reports coming in against him. He was
indicted and the local Chief says: “I have no doubt of his conviction
had he not died since.”
Hardin County, Iowa, had an organization which kept this community
decent and orderly and up to the front in all of the war activities.
The chief was a member of the Bureau of Military Affairs for Hardin
County, which had charge of all the war work. He was also on the
County Committee of Four on Military Instruction, whose duty it was
to instruct and train drafted men. Other members of the A. P. L.
were on the Legal Advisory Board and also were of assistance to the
drafted men. A steady-going and firm-stepping community.
Corning, Iowa, worked in the usual unostentatious way with the
Food and Fuel administrations, etc. Two indictments were brought
against a man who blocked war activities, the fines going to the Red
Cross.
Green County reports: “All quiet in this section. Very few Germans in
our county. None showed disloyalty except one old German woman
who wrote to her son, a missionary in China. Her family promised to
keep her loyal. We examined into the German Lutheran schools and
German language assemblages. Nothing of much consequence.”
Decorah, Iowa, is another peaceful community in a peaceful State.
Little or no trouble was met here. “The A. P. L. was organized rather
late,” says the report, “owing to the fact that we had a most
thorough and efficient Defense Council at work.”
Indianola, Iowa, is also a place of peace. The League had been
organized only a short time when the Armistice broke, and there
were but few activities. “Indianola has a rural population,” says the
Chief, “with a very small percentage of foreign born. No trouble of
any consequence.”
SOUTH DAKOTA.
Aberdeen, South Dakota, must have been a good talking point for
German propagandists, because it reports 122 cases of propaganda
by word of mouth, and 128 cases of propaganda by printed matter.
The division was called on to take active part in the I. W. W. labor
troubles, and this part of its work is described at some length in the
Chief’s report:
Thousands of I. W. W.’s drift here at harvest time. Their
jungles sometimes contain as many as one thousand men.
They take charge of whole trains, and force railroads to
carry them wherever they wish. They have forced the city
authorities in small communities to send them a specified
amount of food, and have defied the authorities of larger
cities to control them. By their methods of sabotage,
murder and arson they have terrorized certain sections of
this state and destroyed millions of dollars’ worth of
property. In the summer of 1917 the annual influx started.
The A. P. L. was called on for assistance, and decidedly
effective measures were adopted. Home Guards and
citizens were organized—later called by a D. J. officer “the
Klu Klux Klan of the Prairies.” Anyhow, this section of the
prairies was soon clear. In consequence, a strike was
declared by the Minneapolis branch of the I. W. W. and
some of their gunmen were sent out. The property of the
Chief of Police at Aberdeen was burnt. In less than two
weeks four of these men were under arrest and two of
them are now serving sentences in the Federal
Penitentiary at Leavenworth. The methods adopted by this
branch of the A. P. L. have proved efficacious. Thousands
of dollars’ worth of property have been saved.
As Aberdeen is located in one of the Non-Partisan League districts,
and as reports have come from nearby towns denoting a large
percentage of pro-Germanism, it may be well to quote further from
the report of this division. The Chief says that one family living in
Hecla, strongly pro-German, declared they would never be taken
alive. The A. P. L. took over the case. One man was shot resisting
arrest. Five members of the family were arrested and two were
convicted, while one remains to be tried. “This stopped pro-German
utterances in that community,” says the Chief, “and materially aided
in the sale of bonds.”
In December, 1917, Fred H—— of Aberdeen was interned for pro-
German utterances. His wife turned state’s evidence on members of
the local German club where members had been fined for speaking
the English language. Four of the leading spirits of this club were
taken into custody, one of them the publisher of three German
language newspapers of wide circulation which were openly pro-
German. This man had sent to von Bernstorff $10,000, ostensibly to
be used for the German Red Cross—all of it raised from readers of
his publication through the sale of the “iron ring.” This man was
sentenced and fined $500. An associate editor of the same string of
papers was interned also. One of the parties was president of the
South Dakota German-American Alliance, and published a German
language paper at Sioux Falls. He was charged with writing a letter
which reads as follows:
I have never given any declaration of loyalty and never
will do it, nor subscribe to any Liberty Loan. The name is
to me already an emetic because hypocritical and
misleading. That a man perhaps buys bonds for business
considerations, I can understand, but I myself couldn’t do
it without thinking that my $50 or $100 might perhaps buy
the explosive which American accomplices of the allied
plunderbund might throw on the house of my mother.
The writer of the above, as head of the German-American Alliance,
raffled a picture of the crew of the Deutschland after our declaration
of war, and sold souvenirs from the boat, remitting the funds to New
York German centers. He was sentenced to ten years in the Federal
penitentiary.
The active Chief of Aberdeen also caught H. M. H——, a former
lieutenant in the German Navy and an ex-instructor in the Naval
School at Hamburg, who was also active in the German-American
Alliance. He got five years in the Federal penitentiary for urging
young men of draft age not to enlist. Another alien enemy whose
papers show that he once had wealthy connections in Germany,
although he was engaged in making a scanty living at baling hay,
was reported as a Prussian and believed to be dangerous. Yet
another, William B——, was picked up in Aberdeen and told a tale
that sounded like one by Deadwood Dick. He said he lived in the
mountains of California with his uncle, who was a smuggler. He was
found to be communicating with the I. W. W., and was sent to a
detention camp. Another arrest was made, of Ed. R——, a wealthy
farmer who stated he would rather see his daughter in a house of
prostitution than a member of the Red Cross. He was sentenced to
five years in the penitentiary, and this has discouraged the
expression of such sentiments near Aberdeen.
Now, if there were nothing else whatever printed in these pages, the
foregoing would show the necessity for such an organization as the
American Protective League, even in communities far away from
manufacturing centers and not supposed to be governed by the
foreign element. The report of the Chief of the Aberdeen Division
affords grave reading and matter for grave consideration. In that
one little community, which does not turn in memoranda of all its
cases, there were 312 Department of Justice cases, 156 War
Department cases, and three Navy Department cases. Seventeen
persons were arrested or interned. Perhaps the most noteworthy of
the recommendations made by the local Chief is this: “It has been
the experience of this branch that the communities reached by the
German language publications have been decidedly disloyal. It is our
opinion that action should be urged upon Congress to discontinue
the foreign language press in America.” These last are words of gold.
They ought to be remembered by every man holding office in the
United States and by every man seeking the suffrages of real
American citizens. The time for mincing matters with these gentry
has gone by.
NORTH DAKOTA
Fargo, North Dakota, hands in a report which varies in one important
particular from those received from neighboring districts. The
division was not making trouble enough for the rampant pro-
Germans in Fargo, so the League turned around and investigated
some of its own officers. None the less, the report tells of a story of
accomplishment, there being 101 disloyalty and sedition cases, 109
cases under the Selective Service Act, and eight cases of enemy
sympathizers who threatened the life of the President.
KANSAS
It will be no surprise to those who know Kansas to learn that this
ultra-progressive, prosperous, energetic State was unswervingly
loyal throughout the war, and had few cases of any kind to report. A
few sentences quoted from the reports of several representative
little towns will serve to show the Kansas war temperature varied
from normal but slightly, if at all.
Oswego, Kansas, reports succinctly: “One hundred percent
patriotism—no aliens.”
White City, Kansas, says: “Ours is a community of loyal citizens. We
spoke to a few about talking too much. Nothing serious.”
Council Grove, Kansas, proved to be a great deal quieter than it used
to be in the days of the Santa Fé trail. The Chief says: “We had a
few pro-German sympathizers whose cases we turned over to the
Department of Justice to investigate.”
NEBRASKA
The A. P. L. Division at Omaha, Nebraska, was organized at a rather
late date, July 1, 1918. The Armistice shattered the activities at a
time when there were three hundred members of the League, each
man ready to do what was asked of him. The Omaha Chief reports
sixty cases of disloyalty and sedition, and several thousand
investigations made in conjunction with D. J. as a result of the
slacker raids, as well as 700 in connection with the Department of
Labor.
The Chief at Hastings, Nebraska, says: “I did not know the work
would be so extensive, or that there would be so much to do. We
have investigated some cases for Omaha, and have done a great
deal of work on draft cases for the state and county boards. We
have been glad to do this work, and I am thankful that I could help
my country this much.”
Callaway, Nebraska, has a grievance: “I had one genuine case of
seditious utterance, but we did not get the evidence. This man was
elected State Senator by the Non-Partisan League. He worked
against the Liberty Bond drive. Fortunately, this year our Senator is
not of his sort politically.”
David City, Nebraska, reports the usual routine work. One pro-
German was taken into custody for making seditious remarks, and
was bound over to the grand jury for trial. The local Chief reports
that his organization is being held intact against any future
emergency.
CHAPTER III
THE STORY OF THE SOUTH
The South is, in its percentage as to population, the finest, cleanest,

truest and most loyal part of the United States to-day. It holds more
of the native born Americans, fewer of the foreign born, and fewer
alien enemies than any like extent of our National possessions. The
only pure-bred American population, sufficiently so to entitle it to a
distinct origin-color of its own on the government census maps, lies
along the crest of the southern Appalachians. There, in parts of
Kentucky, Tennessee, North Carolina, Georgia, Alabama, lower
Virginia, there are Americans who for generations have known no
admixture of any foreign blood. You will find illiteracy there, poverty,
small industrial development. That has come about by reason of a
topography which has left transportation undeveloped. The people
have been held back from the westbound progress of the nation
almost as though caught by the cleats of the great flume through
which poured our early Scotch-Irish, Indian-fighting, wilderness-
conquering ancestry. But it is the finest of gold that those cleats
have caught—a clean-bred, persistent type, of the highest honor, the
highest courage, the highest intellectual quality, the highest physical
qualities. Here and here alone you will find a true American type,
come down with little change from our Colonial days. Would God
that every state in the North and West had these men as the real
inheritors of America, and not the snarling mob of foreigners who in
the last few decades have come to be called American citizens. We
have seen in some part how loyal these last have been, how much
they cared for the flag of America.
The stock of our Highlands has furnished us many strong men, many
of our greatest leaders, our greatest statesmen. Above all, it is fierce
fighting stock. It has been held back by lack of education. These
stark mountaineers are far more illiterate than were their grand-
parents. To-day, in a Cumberland cabin, you may find a Latin
grammar, or a tragedy in the original Greek, of which the owner will
say, “I kaint read none of hit. Grandpap fotched it across the
mountings when he come.” “Across the mountains” lay the Carolinas
and Old Virginia, seats of the most cultured and aristocratic life this
country ever knew, and equal to the best of any land. When we lost
that, we lost the flower of the American civilization. We never shall
replace it. There is no America to-day. There never can be, unless
the seed of the old American stock—never lacking in leaders—one
day shall raise its voice as of old in councils where it will find
hearkening.
The South is a wide country, covering a certain diversity of nature,
but it remains singularly like throughout its borders. Politically it is
still the slave of the color question, whose end no man can see. That
same question restricts the South largely to agriculture. Of late,
Northern money and methods have been reaching out for the raw
wealth of Southern mines and forests, even farming lands. It is in
respect of these later slight changes in the character of the southern
life that the A. P. L. has found its main function there. Had it not
been for imported labor, the A. P. L. would have had no alien and
seditious cases, no propaganda and no disloyalty to report, because
it is absolutely true that our Southern States, which once thought
themselves constitutionally justified in secession, to-day are more
loyal to the American flag man for man, town for town, state for
state, than any or all the remaining states in this Union.
This is true; and yet it is also altogether true that a few Southern
States furnished more cases of desertion or draft evasion than thrice
that number of states in any other portion of the Union, even though
with heavy foreign-born population. How can these two statements
be reconciled?
It is easy, and the level-headed A. P. L. chiefs time and again have
made it plain in their reports. A large percent of the selective service
work had to do with brave young fighting men to whom liberty and
personal freedom made the breath of their nostrils. Many of them
were ignorant—more is the pity. While we have coddled the
treacherous European immigrant, we have forgotten our own
children. Better had we thrown the maudlin Statue of Liberty into
the sea, or turned its face about the other way!
The young Southerner who could not read grandpap’s Latin book, or
any other book, who saw no daily paper and knew nothing of the
outside world, knew only that he did not want to fight in a war of
which he knew nothing and in which he did not think he or his had
any stake. Nobody had threatened him, no men had stolen anything
of his, he did not know where Germany was, and he had never seen
a German to learn to hate him. Why should he fight? He concluded
he would not fight. He would just hide till this war was over, because
it was none of his war.
Very much of the A. P. L. work in the South had to do with getting
into the young man’s comprehension that our Flag was in danger;
that our women and children had been killed by men that did not
fight like men but like brutes. Once that got into the mountain man’s
mind, the day for desertion was past and gone. There are no braver
or more skilled fighting men in the world than in these Southern
hills. There are none more loyal. They did their part and were ready
to do it wherever called. They helped win the war for America as
well as those from richer states. Now that the war is over, let
America forget Europe’s sordid sycophants, the grinning reservists of
the “unbeaten” German Army, and turn attention to these, her own
children—no cuckoo product without an ancestry to claim, who have
no love for this country beyond their love for this country’s easy
money.
MARYLAND
Largely Southern in its population, traditions and political
sympathies, yet Northern in its aggressive spirit and industrial
enterprise, the city of Baltimore perhaps is entitled to be called
“American” more than any other big city on the Atlantic seaboard. It
has always been American, and in this war has only proven anew
what has always been known by those who knew Baltimore. A
hundred years or so ago, in the War of 1812, its citizens fought and
fell gloriously in defense of their city before the British. A beautiful
monument commemorates their heroism. In this war, there was no
city in the country more loyal to our Government and our Allies.
Let it not be thought, however, that the enemy was inactive in
Baltimore. Trouble, active and potential, was present at all times.
That it did not flare up into open destruction was no fault of the
trouble-makers. Like all ports of entry, Baltimore has a considerable
foreign element. Thousands of foreigners were employed in its
shipbuilding plants, on its docks, and in the Bessemer steel works
located near the city. Of pro-Germans and alien enemies there was a
plenty. Many of them, indeed, remembering the landing of the
Deutschland at Baltimore before the war, would have welcomed and
aided a wholesale submarine raid by the enemy—were this possible.
However, this did not come to pass, nor did many other things come
to pass that were justifiably feared. The pro-German, the alien
enemy, the agitator, the Bolshevist were held safe at all times.
Baltimore’s many industries were guarded well. Happily, that industry
which has given her world-wide fame—the oyster industry—required
no protection, and it is a pleasure to record that the nation’s supply
of sea-food was uninterrupted during the war.
A prolific source of trouble for the Baltimore Division lay in the city’s
proximity to the national capital. The overcrowded condition of
Washington during the war forced a huge overflow of population into
Baltimore, and thus doubled the amount of work that otherwise
would probably have been required. This work was tackled with
energy and efficiency by the Baltimore Division, which was one of
the very largest for a city of its size in the country. When the
Armistice came, there were 2,500 operatives engaged in the
multifold activities of the League. The following report does not
begin to tell the full story of their achievement:
Alien enemy cases 110
Sedition and disloyalty 685
Character and loyalty 309
Draft evasion 546
Deserters 225
Liquor and vice 100
Food Administration 3
Miscellaneous 110
Baltimore Division organized and was on the job during the very first
month of the war. Its first Chief was Mr. Edmund Leigh, who solved
the many knotty problems of organization and finance which arose
in the early stages of the League’s growth. Mr. Leigh was succeeded
by Mr. William J. Neale in August, 1918, who acted as head of the
division until November, 1918, when Mr. Tilghman G. Pitts became
Chief.
VIRGINIA
Norfolk, Virginia, was fortunate in having as its chief a gentleman
very prominent in all the war charities, and also of such generosity
of nature that he paid all the expenses of the League out of his own
pocket.
Conditions might have been much worse at this seaport locality, for
only eight cases of alien enemy activity are listed, and five cases of
disloyalty and sedition. This division, however, was able to do a great
deal of work for the War Department, and among other matters
found one illicit still and made four I. W. W. investigations. Another
phase of the work was supplying the M. I. D. officer at the Army
Supply Base—Quartermaster’s Terminal—near Norfolk, with many
photographs of alien enemies and slackers. The Division had
operatives in Army and Navy headquarters, among workmen, etc.,
and had such men included in its personnel as bookkeepers,
timekeepers and others whose work was much appreciated by
Military Intelligence. The chief had twenty-one assistants, all good
men.
White Sulphur Springs, Virginia, had one typical pro-German case.
Adolph S——, a baker of this town, held certain opinions which
would not strictly classify as American. When asked to purchase War
Savings Stamps, he expressed himself as follows: “To hell with your
War Savings Stamps. If Uncle Sam didn’t have money enough to
finance the war, why did he go into it? When the American soldiers
get to France, you’ll find they won’t do anything but run like hell.”
He said a great deal more in similar vein, which “was hardly
suitable,” says the Chief’s report, “for polite ears.” In the U. S.
District Court, at Charleston, S—— confessed to a violation of the
Espionage Act, was fined $100 and sentenced to two years in the
penitentiary.
Lynchburg, Virginia, reports that it was rather quiet. One thing it did
was to draw the fangs of an organization which was formed to
punish such pro-Germans and war obstructionists as the law did not
touch. The A. P. L. has always done its work hand in hand with the
law, and throughout the war has resolutely set its face against
anything savoring of lynch law.
Considerable local trouble arose from returned negro soldiers,
discharged from service, who stated that they had saved the world
from Hun oppression and were entitled to recognition. These
statements had effect on the ignorant population, and it is firmly
believed by the Chief that the “South has a problem on its hands in
this connection which will require considerable time, effort and
patience, if not bloodshed, to solve.” Any one acquainted in the least
degree with the great problem of the South will realize the gravity
and sincerity of this comment.
WEST VIRGINIA
There were “hot times in the old town” of Hinton, West Virginia, in
good part by reason of the activities of one man, the local Chief,
who, for some time was cook, captain and mate of the Nancy brig.
Local disloyalty induced him to go to Washington and ask
government help, and the League organization followed. One pro-
German in Hinton had the Kaiser’s picture on the wall. It is not there
now. The head of this family was a locomotive engineer. The Chief
notified railroad officials not to allow him to handle any troop trains.
Another engineer expressed the belief that a troop train was carrying
“some more fish bait.” He was also relieved of any future work on
troop trains. Two school teachers, after talking with the Chief, hung
up four United States flags and began to sing all the latest war
songs as well as take an active part in Loan drives, Red Cross work,
etc. The largest hotel in the town did not speak well of the war, and
the Chief notified the officers in charge of troop trains to get their
meals somewhere else. A local newspaper printed an article
reflecting on the Red Cross canteen. “I had all the papers publish an
article over my signature,” says the Chief, “that any criticism of the
Red Cross should be addressed to the Bureau of Investigation at
Washington. For this I have been commended by the Red Cross
membership.” It appears that he ought to be commended for his
own record, which, on the face of it, is in the blue-ribbon class.
NORTH CAROLINA
Lexington, N. C., is in the southern mountains. The Chief says:
“Owing to the peculiar reaction of the mountaineer’s philosophy to
the draft laws, many of them ‘stepped back’ into the ‘brush’ to wait
until the war was over. We spent much time in traveling around
among the lumber jacks and sent out word to many delinquents. It
was a simple thing to reach most of these men through the medium
of some trusted friend—much simpler than sending armed men into
the laurel thickets after the fugitives. I don’t believe there is one
case out of ten in western North Carolina where any of our men
avoided the draft through a malicious motive. Whenever a friendly
adviser could reach them to explain the situation, the majority of
them gladly came out. We often made trips of from thirty to fifty
miles into the isolated sections. At one point thirty miles from a
railroad we got information which was sent across the sea to France
and stopped an undesirable appointee to Y. M. C. A. work there.
Some humorous things came up in our mountain travels. One day
our road dwindled to an almost obliterated trail with grass growing
all over it. We sighted an old woman, the first human being seen for
several hours, and asked her if that was the right way to Doeville.
The old woman looked at us with great contempt, and remarked:
‘Lord bless us, you-all is right in Doeville dis minute!’”
The Chief of Lexington says that not everyone understands the
mountain boys and that they certainly make excellent fighters when
in the army. “One of them in my district,” reports the Chief, “had to
be run down and captured by his own father, who delivered him over
to the authorities for military service. This boy was the first of his
company to distinguish himself in France.”
The Chief of Salisbury, North Carolina, Division sends in his final
report in homely and convincing phrases, a mark of the good
common sense employed in his work. One pro-German was called
into the office and the Chief said to him: “Mr. ——, I hear that the
next time you and your family come to town over the public road,
you are going to be blown up without any warning.” The man struck
the table with his fist and said: “I’d like to know how! The public
road is mine and I’m going to travel on it.” The Chief said: “So our
ships had a public highway to Europe. The Germans have destroyed
vessels, women and children without warning. What do you think of
it?” The pro-German thought this over a minute and exclaimed:
“Why hasn’t some one talked to me like that before? I never saw it
that way before.”
Hickory, N. C., says: “Our work was largely educational. We had no
aliens—all native born American citizens. Thirty of our leading
citizens constituted the membership of the League. When we went
to work, all the ’aginners’ who were against the war got on the right
side. Especially was this true after the amended espionage act went
into effect. In my judgment,” says the Chief, “the psychological effect
of an organization that could be felt but not seen helped wonderfully
in bringing to their right senses the small minority that were not in
right at the start.”
Durham, N. C., pulled off one raid on a circus crowd and got ten
slackers. “Our community has a foreign element,” says the Chief,
“and is above the average in respect to law and order. Our members
were prominent in the war activities.”
SOUTH CAROLINA
Anderson, S. C., says: “Our organization has been anxious to answer
every call. There are practically no foreigners in this section, so
violations of the war measures have been almost negligible. Most of
our work has been making reports for overseas service. The men all
consider it a great honor to have been members of the League.”
A man whom we may call Benny Vogel deserted from the 105th
Infantry at Camp Wadsworth, South Carolina. In some way, he found
his way to Schenectady, New York, where he proceeded casually to
marry a young lady of that city, under date of April 19, 1918. The
wife was watched. The deserter was caught and returned for
punishment.
St. Matthews, S. C., reports: “On the whole there was little enemy
activity. We unearthed six cases of discharged soldiers drawing
government money who were not entitled to it, and eight cases of
parties receiving allotments from soldiers for incorrect amounts. We
changed such undesirable sentiment as existed in our community,
and with tact and judgment rather than by drastic measures. We
think our community is among the most loyal of any in America and
doubt seriously if there is one per cent disloyalty here. Some who at
first were lukewarm changed, and we knew it was due to the policy
adopted by our organization. We worked on the Sunday law and the
fuel laws, the food regulations, etc., all in a quiet way, but, we think,
with good results throughout our county.”
GEORGIA
All sorts of stories show in the League files. One regarding
submarine bases along the Georgia and Carolina coast was traced
down to the purchase of a piece of land by a former grocery clerk, a
naturalized German, who resided in Savannah for many years. He
was outspoken in his sympathy with Germany before the United
States entered the war. A report made by the Navy Department to
the National Directors of the League states:
“On January 6, 1918, this man was tried in the city court of
Savannah and found guilty of violating the prohibition laws. He was
fined $400 and sentenced to six months on the chain gang. Before
he had fully served his sentence he was re-arrested by the United
States Marshal on a presidential warrant and subsequently interned.”
The brief phrase “presidential warrant” covered many and many a
case of naturalized Germans who became too loquacious in this
country before and after we entered the war.
Atlanta, Georgia, had a nice scare about the report that a German U-
boat captain had landed and was on his way to Atlanta, dressed in
an American officer’s uniform. Operatives were out and trailed every
military or quasi-military looking man on the streets or anywhere
else. Their first haul included a major from the Judge Advocate
General’s office and a Judge from the Federal Court. The next alarm
came from two operatives who trailed an officer just off the train,
who turned out to be a colonel of the Quartermaster’s Corps,
U. S. A. The latter was able to make his escape. The Chief adds:
“Just how many suspects were held up that night it would be difficult
to state. Operative No. 3 turned in a report of his activities the next
morning. It seemed he had held up the following personnel: One
Lieutenant-Colonel, sixteen Majors, twenty-three Captains, forty-two
Lieutenants, one Lieutenant-Commander, three Ensigns, and seven
Sergeants—a total of ninety-two suspects.” He closed his report with
the following heartfelt remarks: “Well, I didn’t know what kind of
uniform the German had. Besides, every man I stopped was a blond.
I didn’t stop any other sort.” D. J. reported it was satisfied that no
German submarine officer had visited Atlanta.
ALABAMA
Birmingham, Alabama, was one of the most active and interesting
divisions of the League. It took on 1,849 cases under the Selective
Service Act, 76 investigations of pro-Germans, 123 cases of
deserters, and 153 Red Cross loyalty reports, besides a large list of
general war activities. Some of the star cases of deserter hunting at
Birmingham are reported in another chapter.
Like many another community, Birmingham also had its wireless
case, and like most cases of the sort throughout the country, it
created much excitement in the division while it lasted. Certain
mysterious light flashes, supposed to be signals, were reported
along the top of a high hill on the outskirts of the city. Operatives
detailed on the case could learn nothing, but still reports kept
coming in. Finally, one astute visiting chief followed a high-powered
transmission line along the mountain and found that the limb of a
tree at a certain spot would touch the wire when swayed by the
wind. The repeated rubbing had worn away the insulation, exposing
the bare wire. When the limb came in contact with the wire,
especially during a rainy night, a spark would be made when the
limb and wire separated: The Chief adds: “When the limb was cut
off, we received no further reports of mysterious signals.” There
have been bluish-white lights which some thought indicated a
wireless outfit in operation.
Montgomery, Alabama, reports one of those curious cases which
were sometimes met with in the course of the League’s
investigations. This was a straight-goods, dyed-in-the-wool, bona-
fide conscientious objector. His name was W. A. P——, a farmer who
had a son in the draft, but who needed him on the farm. He
accompanied the boy to the examination board, after the young man
had been arrested by the sheriff. He brought his Bible to the board
and tried to prove that he was justified in his objections; that he was
responsible for the care of this boy; that the Lord had given him that
duty and no one else. The old man was violently opposed to
bloodshed and quoted the scriptural words, “Thou shalt not kill,” and
“Children, obey your parents.” The Chief had a long talk with him at
his farm. He admitted that he told his son not to answer questions,
and that he had another son who had attained his eighteenth
birthday and had not registered. The Chief told him to be careful or
he would get into trouble. He said, “I am not getting into any
trouble; it is you people who are provoking the wrath of God.” All the
agent could do was to tell him that he must come before the United
States Commissioners. P—— was brought in to the Committee, and
bound over to the grand jury. Before the trial, he stood up and
remarked, “Let us have a word of prayer,” and prayed fervently for
several minutes. He carried his Bible with him at all times. P——
seemed to be generous. “He came to Montgomery and brought a
couple of gallons of nice syrup for the Deputy and Commissioners,”
says the Chief. One would think that the A. P. L. would be glad to
have peace at any price in such surroundings, even without syrup.
Selma, Alabama, is another one of the loyal Southern communities.
“We kept down seditious utterances,” says the Chief. “Without doubt
we have had a most wholesome effect on our citizenry by letting
every one know that this was not a time for anything that was not
one hundred per cent American. I do not believe there was a greater
force for good in the State of Alabama than the American Protective
League.”
FLORIDA
Cocoa, Florida, is not far from one of the Government shipyards, and
so had had some contact with persons inclined to be pro-German.
By way of explaining the additional activities sometimes taken on by
the League, the Chief says: “This office worked with the Special
Agents at Jacksonville, and with officers of the Seventh Naval
District. We have also given information to the Collector of Internal
Revenue concerning those who should pay income tax. Our division
consisted of twenty-four members—all high-class men who could be
relied upon in any emergency that might arise. We were taking steps
to enlarge the organization when the German balloon burst.”
Eustis, Florida, was more especially concerned with war cases. Forty-
one cases of draft delinquency were handled; two slacker raids were
conducted, and there was a little “work or fight” activity. Eustis is in
a county which had the reputation of harboring a good many
slackers and deserters, who sought peace and quiet in some of the
out-of-the-way places. Through the activities of the local A. P. L.
division, this situation was cleared up distinctly. The Chief says: “We
believe we have been instrumental in protecting many people from
their own follies, and have brought to justice men who were
engaged in obstructing the Government’s war activities in one part
of the country or another. It has been a pleasurable though arduous
service that some of us have rendered in this work.”
Kissimmee, Florida, reports: “All quiet along the Kissimmee. Our
community was singularly free of annoyance of any character. Two
or three persons were indiscreet in their language, but we found
that a small reminder was sufficient to stop the talk.”
KENTUCKY
Louisville, Kentucky, is a busy and famous old town with a reputation
for being engaged in the manufacture of trouble-making products,
but there seems to have been very little trouble. Only eighty-nine
cases of disloyalty and sedition are reported, and 308 under the
selective service regulations.
Mr. George T. Ragsdale, the first Chief of Louisville Division,
instructed his men to keep under cover, so that the personnel of the
division was very little known. More than 700 reports were made in
all, and nine men were sent to the penitentiary. Local business men
furnished most of the working capital. Upon Mr. Ragsdale’s
resignation, Mr. J. V. Norman was appointed Chief, taking over about
400 members. The city was divided into nine districts and the
County in three, with the usual subdivisions of captains and
lieutenants as operatives. The membership was up to about 700 at
the time of the signing of the Armistice.
Most of the investigations handled by the Louisville Division were on
requests coming from local draft boards, although the several
branches of the government’s legal organization frequently asked for
aid. Several thousand men were questioned in the slacker raid of
August 3. Thirty-five men were taken to jail and fourteen inducted;
among these, several deserters. Sometimes at a race track a quiet
investigation would be put on without any open raid.
Among the list of delinquents turned in was a man named Lyle D. B
——. An intercepted letter resulted in an examination of the man’s
mother, who refused to tell where he was. Portland, Oregon, was
suspected as his present residence. The case came to an end when
it was found that the delinquent had been committed to the Federal
penitentiary at McNeil Island, Washington. His questionnaire was
forwarded by the local board to the penitentiary and returned
properly filled in. The man had a fairly good alibi. The usual cases of
religious fanatics, loud talkers and bearers of false witness were
uncovered in the League’s work. Many of the best citizens of
Louisville were engaged in these somewhat undignified and often
thankless tasks of ferreting out such matters.
Lexington, Kentucky, as might easily be expected, reports in
American fashion: “The sentiment of our entire population is hard
against the Germans and their allies. Our people are almost
unanimous in their opposition to showing Germany any
consideration, even with furnishing them food after their defeat. The
one sentiment is that Germany could feed herself while in war; now
let her feed herself since she is out of war.”
The work of the Lexington Division was mostly concerned with the
local and district boards. It handled 405 cases of this sort. There
were only thirty cases of disloyalty and sedition investigated, and
forty cases of word-of-mouth propaganda.
Marion, Kentucky, says: “We are glad to report that our county has
been so patriotic that little of any importance is required to be done.
We had to caution a few of our citizens as to the bad results of
opposition to the United States in the war. We have no foreign
element. Our citizens come from Virginia, Tennessee and North
Carolina, and are of old families. We rarely see anyone of foreign
descent in this section except traveling men who make trips through
the county.”
Somerset, Kentucky, had a bad man—a deserter who escaped from
Fort Oglethorpe once or twice, the last time taking along his rifle and
pistol. He barricaded himself in an old house at Helenwood,
Tennessee. The A. P. L. took him all right, in spite of his threats. He
is in Fort Leavenworth for twenty years. From far off Livingston,
Montana, came a request to Somerset Division to arrest one Willie
McK——, a professional evader. He was found attending church. The
Chief says: “We walked in and gave him a tap on the shoulder, and
told him to come out. Just as we started for the door, the choir
began to sing, ‘God be with you till we meet again.’ It is going to be
some time.” Somerset concludes: “We did not stop when the
Armistice was signed, but kept watching everything and giving the
Government the best that was in us!” Isn’t that fine?
TENNESSEE
The A. P. L. work in the beautiful and historic old city of Nashville
was somewhat circumscribed because of the activities of other
agencies already in existence. The division did its share in the
routine work of war activities, apprehending evaders, conducting
numerous investigations, and vigilantly keeping tab on the comings
and goings in the Old Hickory Powder Plant.
Chattanooga, Tennessee, did its bit and did it well. Ten prisoners
who escaped from the local War Prison were apprehended by
division operatives, and brought back for reinternment. One member
of the division discovered an extensive system of graft in connection
with the Government construction work on the Nitrate Plant at
Sheffield, Ala. Report of this was furnished to a Special Agent, who
was detailed by the Government to conduct an investigation. The
Chief comments: “Just what can be proven in this case remains to
be seen.”
Some of the most amusing Chattanooga investigations were those of
the religious sect known as the “Holy Rollers.” Several of these
preachers had preached sermons in which they condemned the Red
Cross and the Government generally. These men were apprehended,
and members of their congregations testified at local headquarters.
Some of these preachers were moved by the “spirit” in their
testimony, but after they remained in jail a short time, they saw the
Scriptures in a different light, and very few of them offended a
second time.
Another Chattanooga case had in it the possibilities of great
mischief. A large amount of mail to an illiterate mountaineer caused
an A. P. L. operative and a Special Agent of the Department of
Justice to go to the top of Sand Mountain, and in a dirty log cabin
they found a wagon load of I. W. W. literature and correspondence
in which were letters from Emma Goldman and other leading lights
of socialistic faith. The man himself was working in a foundry turning
out Government orders; he was organizing a strike at the time he
was taken into custody.
Clarksville, Tennessee, is in the loyal Southern country, and is very
free from alien population. There were only twenty-five
investigations for disloyalty and sedition, and propaganda was
almost negligible. As this is the tobacco producing section, there was
considerable property investigated under the Trading with the Enemy
Act, and some helpful reports were made to the Alien Property
Custodian. The League members were active in all the war work.
Hopkinsville, Tennessee, had a great deal of trouble over illegal
transportation of whiskey, a great deal of which went to workers in
government powder plants in an adjoining city. “We arrested so
many that no record was kept,” says the Chief. Things became
quieter later on.
Huntingdon, Tennessee, is another disgustingly quiet and satisfied
community. “People nearly all natives,” says the report, “and mighty
few expressions of disloyalty. We have watched for violations, but
nothing has developed worthy of report.”
TEXAS
San Antonio, Texas, is in a strongly pro-German neighborhood and
has a large citizenry of German descent. It is refreshing nevertheless
to see that in this good old Texas town, once distinctly Spanish, the
language of the United States prevails to-day and only one flag
floats over the Alamo. There were thirty-four investigations for
sedition, and twenty-four cases of propaganda. The usual number of
overseas examinations were held. On the whole, San Antonio seems
to have been quiet and peaceful and distinctly loyal in every way, in
spite of her location so close to New Braunfels.
The San Antonio Chief concludes his too brief report with a little
story:
The telephone at my elbow rang insistently. The man at
the other end of the wire was incoherent, and I could not
understand what he wanted.
“Hold on a minute!” I finally interrupted. “Who is this
speaking?”
He would not tell me; he merely said that he was a friend
of mine. I did not like to give information over the ’phone
when I was not sure as to whom I was talking. I again
insisted that he give me his name; once more he refused
to do so, reiterating that he was a good friend of mine. I
could not recognize the voice. But what he said was
startling.
Recently I had been appointed Chief of the American
Protective League for this District, and how my informant
had learned, or guessed, that I was engaged in it, I could
not tell. I did not like to undertake a wild goose chase; at
the same time, if I should refuse to follow up the clue he
gave me, the lives of many might be endangered.
Anything could happen in San Antonio. It is one of the
oldest cities in the United States, and ever since the day
the Spaniard founded it, has been a hotbed of intrigue.
Just at this time there were fully twenty thousand troops
stationed in the various Camps about the City, and in
order to impress the Mexicans with the idea that we were
not altogether helpless, it had been suggested that a
patriotic military parade be given. This was to take place
the following day, and I had spent many hours helping to
arrange the details. And now, my mysterious “friend” had
told me over the ’phone that he knew certain parties were
plotting to throw a bomb into the parade; that if I would
go to the certain house named by him, I would find a
meeting of the plotters in progress!
There was no time to be wasted. I got in touch with one
of my lieutenants, M——, and asked him to meet me in
half an hour, and to come armed. Before leaving the office
I sent for a couple of suits of overalls, one of which I
donned, and when I met M——, I gave him the other.
I told him all that I knew, and he realized that it was
serious. We parked our car about two blocks from the
house designated by my informant, and approached it
afoot. The neighborhood was questionable. The house to
which I had been directed stood a few feet back from the
street in a neglected tangle of shrubbery. There was a
fence about the property, but no gate. It was a small
frame shack with two rooms in front and a third forming
an ell. We walked around it cautiously several times, and
finally discovered a light in the ell. The blinds were all
tightly closed, and it was but a faint glimmer through a
crack that we saw. We crawled carefully to the gallery and
each looked through the crack.
We could barely distinguish the forms of five men huddled
over an oil stove in the middle of the room. Three were in
overalls and had the appearance of laborers; one wore a
shabby old suit of civilian clothes, and the fifth appeared
to be in uniform. Their heads were close together and
they seemed to be talking in low tones, but neither M——
nor I could distinguish a word that was said.
There was a door a few feet from where we were, and I
noticed another one on the opposite side of the room. I
told M—— to go around to the other door and I would
remain where I was. If either of us was able to distinguish
any suspicious words, or if we found any reason to
suspect that the five men were actually plotting, a low
whistle was to be the signal to the other, and
simultaneously we were to break in the door and rush
them.
While the whole thing had the appearance of a conspiracy,
and I was inclined to take the bull by the horns and give
M—— the agreed signal, I was also suspicious that
someone might be playing a practical joke on me. While I
hesitated, M—— suddenly sneezed!
I have lived in the Southwest the greater part of my life
and have been in some pretty tight places, and always
have prided myself on my ability to take care of myself in
an emergency; but the next thing I knew after M——’s
sneeze, he was bending over me trying to staunch the
blood that was flowing from a wound over my right eye,
at the same time reading the riot act to me in choice
language.
“What happened?” I asked, feebly.
“Why, the whole darned shooting-match jumped your way,
walked over you and beat it!” he explained in
exasperation. “What I’ve been trying to find out is why in
hell you didn’t shoot?”
I could not answer in words, but mutely I showed him
that in my haste I carefully had put on the overalls over
my clothes with my gun in the usual place in my hip
pocket. It would have taken me five minutes to get it out.
“It’s a good thing you had it so well hid,” he remarked.
“They might have taken it away from you!”
We searched the deserted house. Except for the stove it
was devoid of furniture, and we found nothing in the way
of a clue.
We arranged for a strict patrol of the route of the parade.
Each man was given a “beat.” If any man saw anything
suspicious, and particularly a suspicious package, he was
to investigate and report at once.
The parade was crossing the Houston Street bridge,
where I happened to be, when I saw a negro man
elbowing his way to the front of the crowd along the curb.
In his right hand, held high over the heads of those about
him, was a package wrapped in newspaper! He seemed in
the act of hurling it into the street when I sprang forward
and grabbed the upraised arm, dragging the negro back
to the railing of the bridge.
“What have you got in that package?” I demanded.
“My Gawd, boss, you’se the fou’th man to ast me about
ma lunch in the last five minutes. If it’s worrying you
white folks so much, guess I’d better git shet of it!”
Before I could prevent him, he threw it into the river, and
turned to view the parade with a muttered opinion on my
interference with his personal liberties. All we succeeded
in accomplishing was scaring a poor negro out of his
lunch, but whether or not we thwarted others in a worse
plot, we never knew.
But that was much our story in San Antonio. We did the
best we knew. Had we not been there, and were it not
known that we were there, matters might have been
worse. The makings of trouble were around us all the
time.
Laredo, Texas, on the Mexican border, was organized for business.
The Chief says: “We have very few alien enemies resident here.
Before we organized, there was some talk of a disloyal nature, but
this situation changed at once when it got out that we had seventy-
five or eighty members whose identity was unknown to the public
but who would be pretty sure to be out for business. For the six or
eight months before the Armistice we heard scarcely a word
unfavorable to the United States or her Allies. We think we did
something in the way of prevention if not of cure.”
Yoakum, Texas, has ten cases of disloyalty and a like number of
word-of-mouth propaganda. A good local chief of a fighting family
says: “We were ready at all times to meet any emergency regardless
of distance or difficulty.”
Beaumont, Texas, is in the oil country, and such centers quite often
attract alien population. The Beaumont report covers sixty-three
cases of alien enemy activities, eighteen cases of disloyalty, and
ninety cases under the selective service regulations.
ARKANSAS
Cotter, Arkansas, reports that it is a community with very few
foreigners, the population being American for generations back. The
Chief says: “We had two deserters who lived for two weeks in an
inaccessible camp in the mountains. They finally got hungry, came in
and surrendered. We also had one draft-dodging case of a peculiar
sort. This young man, according to his marriage license, should have
registered in June, 1917. He did not. We traced him to Oklahoma,
and from there to Springfield, Missouri. He was taken into custody
by the Chief of Police at that point on our order. We sent a certified
copy of his marriage license, but he had enough of his relatives on
hand to swear to his true age, to secure his release.”
Helena, Arkansas, also comes into court with very clean hands. Its
report shows a membership of 127, which proved to be none too
large, as all hands found work to do. Investigations were handled all
over Arkansas, Mississippi and Louisiana.
Fort Smith, Arkansas, found its slacker raids more interesting than
anything else. It conducted two of them, a slacker or two being
apprehended each time. One stranger, who was sufficiently
indiscreet as to fail to register, was unceremoniously hauled out of
bed and turned over to the local war board. No alien enemy
activities came to the attention of this division.

[Ebooks PDF] download Machine Learning for Practical Decision Making: A Multidisciplinary Perspective with Applications from Healthcare, Engineering and Business Analytics Christo El Morr full chapters

Uploaded by

Copyright:

Available Formats

[Ebooks PDF] download Machine Learning for Practical Decision Making: A Multidisciplinary Perspective with Applications from Healthcare, Engineering and Business Analytics Christo El Morr full chapters

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

[Ebooks PDF] download Machine Learning for Practical Decision Making: A Multidisciplinary Perspective with Applications from Healthcare, Engineering and Business Analytics Christo El Morr full chapters

Uploaded by

Copyright:

Available Formats

Full download ebooks at https://ebookmeta.

Machine Learning for Practical Decision Making:

Machine Learning for Business Analytics: Concepts,

Machine Learning for Business Analytics: Concepts,

Business Analytics A Data Driven Decision Making

Deep Learning Applications and Intelligent Decision

Practical AI for Healthcare Professionals: Machine

AI and Society Tensions and Opportunities Chapman Hall

Developing Informed Intuition for Decision Making Data

Spreadsheet Modeling and Decision Analysis: A Practical

Editorial Board Members

Hossam Ali-Hassan Walid El-Hallak

ISSN 0884-8289 ISSN 2214-7934 (electronic)

Toronto, ON, Canada Christo El Morr

1 Introduction to Machine Learning . . . . . . . . . . . . . . . . . . . . . . . . . 1

3.11.2 Supervised Learning Overview . . . . . . . . . . . . . . . . 98

4.12.2 Working Example in Weka . . . . . . . . . . . . . . . . . . . 150

7 Logistic Regression . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 231

9.3.2 Gaussian Naïve Bayes Example . . . . . . . . . . . . . . . . 285

11.3.9 Backpropagation . . . . . . . . . . . . . . . . . . . . . . . . . . . 342

The South is, in its percentage as to population, the finest, cleanest,

You might also like