Scribd Logo
Upload

Welcome to Scribd!

  • 432 views

    Shark Tank - Web and Social Media Analytics Case Study

    Uploaded by

    Shyam Kishore Tripathi
    This document summarizes a project using predictive modeling on a Shark Tank dataset to predict whether investments were made based on entrepreneurs' pitches. The following steps were taken: data was cleaned and preprocessed; CART, logistic regression, and random forest models were built using pitch descriptions alone and with an additional funding ratio variable; models were evaluated and compared, with logistic regression achieving 100% accuracy and performing best. Logistic regression was selected as the best model for further analysis.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd

    Shark Tank - Web and Social Media Analytics Case Study

    Uploaded by

    Shyam Kishore Tripathi
    432 views9 pages
    This document summarizes a project using predictive modeling on a Shark Tank dataset to predict whether investments were made based on entrepreneurs' pitches. The following steps were taken: data was cleaned and preprocessed; CART, logistic regression, and random forest models were built using pitch descriptions alone and with an additional funding ratio variable; models were evaluated and compared, with logistic regression achieving 100% accuracy and performing best. Logistic regression was selected as the best model for further analysis.

    Copyright

    © © All Rights Reserved

    Available Formats

    PDF, TXT or read online from Scribd

    Did you find this document useful?

    Is this content inappropriate?

    This document summarizes a project using predictive modeling on a Shark Tank dataset to predict whether investments were made based on entrepreneurs' pitches. The following steps were taken: data was cleaned and preprocessed; CART, logistic regression, and random forest models were built using pitch descriptions alone and with an additional funding ratio variable; models were evaluated and compared, with logistic regression achieving 100% accuracy and performing best. Logistic regression was selected as the best model for further analysis.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    432 views9 pages

    Shark Tank - Web and Social Media Analytics Case Study

    Uploaded by

    Shyam Kishore Tripathi
    This document summarizes a project using predictive modeling on a Shark Tank dataset to predict whether investments were made based on entrepreneurs' pitches. The following steps were taken: data was cleaned and preprocessed; CART, logistic regression, and random forest models were built using pitch descriptions alone and with an additional funding ratio variable; models were evaluated and compared, with logistic regression achieving 100% accuracy and performing best. Logistic regression was selected as the best model for further analysis.

    Copyright:

    © All Rights Reserved
    Download as PDF, TXT or read online from Scribd
    Download as pdf or txt
    of 9

    A Project on Web and Social

    Media Analytics

    SHYAM KISHORE TRIPATHI


    PGP - BABI

    0|Page
    Table of Contents

    Sl. No. Description Page No.


    1 Project Objective 2
    2 Defining Business Problem 2
    3 Reading data and performing initial data clean up 2
    4 CART, Logistic Regression and Random Forest Before Ratio 4
    5 CART, Logistic Regression and Random Forest After Ratio 6
    6 Comparing Model and Conclusion 8

    1|Page
    Project Objective
    In this exercise we will be developing a predictive model to predict deal or no deal using Shark Tank dataset (US based
    show). Complete exercise will be based upon the text mining analytics.

    Defining Business Problem


    Shark Tank is an US based show wherein entrepreneurs and founders pitch their businesses in front of investors who
    decides to invest or not in the businesses based on multiple parameters.

    In this dataset, we have got a dataset containing Shark Tank episodes with 495 records where each entrepreneur
    made their pitch to investors . Using Social Media Analytics algorithms, we will predict whether given description of
    pitch it will convert into success or not.

    Reading data and performing initial data clean up

    1. Loading dataset
    Sharktank = read.csv("Shark Tank.csv", stringsAsFactors=FALSE)

    2. Loading required libraries


    library(wordcloud)
    library(tm)
    library(SnowballC)
    library(rpart)
    library(rpart.plot)

    3. Performing Initial Clean Up


    # Creating corpus
    corpus = Corpus(VectorSource(Sharktank$description))

    # Convert to lower-case
    corpus = tm_map(corpus, tolower)

    # Remove punctuation
    corpus = tm_map(corpus, removePunctuation)

    # Word cloud before removing stopwords


    wordcloud(corpus,colors=rainbow(7),max.words=100)

    2|Page
    Now we need to normalize the texts before we proceed with further analysis. Below are the steps we need
    to perform
    1. Converting every text to lower case
    2. Removing punctuation marks and stop words.
    3. Removing extra white spaces.
    4. Perform Stemming of documents

    We need to use DTM (Document-Term Matrix) for further analysis when basically we will be converting all
    the documents as rows, terms/words as columns, frequency of the term in the document.

    This will help us identify unique words in the corpus used frequently.

    To reduce the dimensions in DTM, we will remove less frequent words using remove Sparse Terms and
    sparsity less than 0.995

    3|Page
    Converting this dataset into data frame and add dependent variable “deal” into data frame as final step for
    data preparation.

    4. CART, Logistic Regression and Random Forest Before Ratio

    To predict whether investors will invest in the businesses we will use deal as an output variable and use the
    CART, logistic regression and random forest models to measure the performance and accuracy of the model.

    a) Building CART Model

    4|Page
    Evaluating CART Model

    b) Building Random Forest Model

    5|Page
    c) Building Logistic Regression Model

    5. CART, Logistic Regression and Random Forest After Ratio

    Now we will add additional variable “ Ratio “which will be derived using column ask for/valuation and then
    we will re-run the models to see if we can have improved accuracy in the models

    a) Building CART Model

    6|Page
    b) Building Random Forest Model

    c) Building Logistic Regression Model

    7|Page
    6. Comparing Model and Conclusion

    Action CART Model Logistic Regression Model Random Forest Model


    Before Ratio Accuracy 65.65% 99.79% 55.35%
    After Ratio Accuracy 66.06% 100% 55.75%

    With CART Model we were able to predict around 65.65% and 66.06% accurate results using only description
    and description + ratio respectively.

    With Random Forest, we were able to predict 55.35% and 55.75% accurate results using only description and
    description + ratio respectively.

    With Logistic regression, it gave us 100% accuracy description and description + ratio .

    From the above analysis we can confirm that Logistic Regression is the best model for proceeding further with
    insight analysis.

    8|Page

    You might also like