Introduction to Machine Learning

Anubha Gupta, PhD.

Professor
SBILab, Dept. of ECE,
IIIT-Delhi, India
Contact: [email protected]; Lab: http://sbilab.iiitd.edu.in
 Machine Learning in Hindi

Introduction to Machine Learning

2
 Machine Learning in Hindi

Motivation
Some real-world examples of ML around us…

Traffic Policing & Challan

https://www.91wheels.com/news/indias-first-ai-enabled-
cameras-installed-on-mumbai-pune-expressway

AI-powered chatbot for customer query handling

https://timesofindia.indiatimes.com/business/india-business/hdfc-bank-
launches-artificial-intelligence-driven-chatbot-
eva/articleshow/57481168.cms

Sana, a virtual news anchor (Aaj Tak)

https://www.jivanhindi.in/2023/04/ai-anchor-sana-kaun- 3
hai-in-hindi.html
 Machine Learning in Hindi

Learning Objectives
● Understand Machine Learning (ML)
○ Past & present
○ Definition
○ Types
● Supervised ML
○ Definition, example, applications, challenges
● Unsupervised ML
○ Definition, example, applications, challenges
● Applications
● Practical Aspects

4
 Machine Learning in Hindi

Evolution of AI over time

Artificial Intelligence (AI)
The ability of
machines to Machine Learning (ML)
simulate human
intelligence, Involves training
such as learning, algorithms to gain
reasoning, and intelligence using data, Deep Learning (DL)
decision-making without being explicitly
Involves training algorithms that mimic
programmed
human brain to gain intelligence using data,
without being explicitly programmed

1950 1980 2000 2010 2020

ML is a subset of AI and a power set of DL.

https://blogs.nvidia.com/blog/2016/07/29/whats-difference-artificial-intelligence-machine-learning-deep-learning-ai/ 5
 Machine Learning in Hindi

History of Machine Learning

● Arthur Samuel, an American AI-pioneer, is credited
to popularize the term Machine Learning (1959) [1],
who explained ML as,

“Field of study that give computers the ability to learn

without being explicitly programmed”

● The 1980s
○ emergence of neural networks (NN) technique
○ NN limitations
○ Hindered to widespread adoption: data
sparsity & low computing capabilities
Samuel’s checkers-playing program was
○ First AI-Winter (10.1109/MIS.2008.20) among the world's first successful self-
learning programs.
https://blog.hnf.de/als-ein-computer-dame-lernte/

[1] A. L. Samuel, "Some Studies in Machine Learning Using the Game of Checkers," in IBM Journal of Research and Development, vol. 3, no. 3, pp.
210-229, July 1959, doi: 10.1147/rd.33.0210.
6
 Machine Learning in Hindi

Machine Learning Today

● The 2000s
○ development of support vector machines
(SVM) & decision trees
■ able to handle larger datasets
■ more computationally efficient [2]
○ rise of big data
○ development of deep learning (DL)
● Today,
○ ML’s wide-spread applications - from image The Three Godfathers of AI: Yann LeCun,
recognition and natural language processing Geoffrey Hinton, and Yoshua Bengio are
(NLP) to self-driving cars and personalized known for their pioneering contributions in
medicine. the development of AI https://fortune.com/2019/04/02/eye-
on-ai-godfathers-deep-learning/
■ DeepFace: DL-based Facial Recognition
■ AlphaGo: AI-player of ‘Go’ board game
[defeated world champion]
■ ChatGPT: AI-chatbot, and so on…

[2] Hastie, Trevor, et al. The elements of statistical learning: data mining, inference, and prediction. Vol. 2. New York: springer, 2009.
7
 Machine Learning in Hindi

Definition
“A computer program is said to learn from experience E with respect to some class of tasks T and
performance measure P if its performance at tasks in T, as measured by P, improves with
experience E [3].”

Utilize Measure
T

E P
Learn

8
[3] Machine Learning, Tom M. Mitchell, McGraw Hill, 1997.
 Machine Learning in Hindi

Types
● Supervised Learning
● Unsupervised Learning

Supervised Learning Unsupervised Learning

9
 Machine Learning in Hindi

Supervised Machine Learning

● Algorithm learns to map input data (features) to known output data (labels).
● Training data is labeled: each data point has a corresponding output label that the algorithm is
trying to predict.
● Tasks:
○ classification: the output variable is a categorical variable (response is qualitative)
○ regression: the output variable is a continuous variable (response is quantitative)
● Example:
○ Classification task of predicting fruits

Model
Training
Prediction
Cherry

Mango

Cherry Mango
10
Training Testing
 Machine Learning in Hindi

Supervised Machine Learning

● Some Applications:
○ Image classification: Identify objects or features in images, such as classifying whether
an image contains a cat or a dog.
○ Sentiment analysis: Determine the sentiment or opinion expressed in text data, such as
classifying whether a product review is positive or negative.
○ Fraud detection: Identify fraudulent transactions in financial data, such as predicting
whether a credit card transaction is likely to be fraudulent.
○ Predictive maintenance: Predict when maintenance is needed for equipment based on
sensor data, such as predicting when a machine is likely to fail.
○ Personalized recommendations: Recommend products or content to users based on their
past behavior and preferences, such as suggesting movies or TV shows to watch.

11
 Machine Learning in Hindi

Supervised Machine Learning

● Challenges:
○ Insufficient or biased data: Lack of data or biased data can lead to poor model
performance or inaccurate predictions.
○ Overfitting: When the model is too complex, it can fit the training data too closely,
leading to poor generalization performance on new data.
○ Feature engineering: Choosing and engineering the right set of features can be a time-
consuming and subjective process that requires domain expertise.
○ Model selection: Choosing the right model and hyperparameters for a given problem
can be challenging and require extensive experimentation and tuning.
○ Interpretability: Understanding why a model makes certain predictions or decisions can
be difficult, especially for complex models like deep neural networks.

12
 Machine Learning in Hindi

Unsupervised Machine Learning

● Algorithm learns patterns and structures in the input data (features) without being given
explicit labels / targets.
● Training data is unlabeled.
● Tasks:
○ clustering: group similar data points together
○ dimensionality reduction: the output represents the input with reduced dimensions
● Example:
○ Clustering task of grouping fruits

Class 1

Class 2
Model
Model Output
Training
13
 Machine Learning in Hindi

Unsupervised Machine Learning

● Applications:
○ Anomaly detection: Identify unusual or rare events or patterns in data, such as detecting
fraudulent transactions.
○ Clustering: Group similar data points together based on similarity or distance, such as
clustering customers based on their purchasing behavior.
○ Dimensionality reduction: Reduce the number of features or variables in high-
dimensional data while preserving as much information as possible, such as reducing the
number of dimensions in an image or text dataset.
○ Topic modeling: Discover underlying topics or themes in a collection of documents or
text data, such as identifying topics in customer reviews or news articles.

14
 Machine Learning in Hindi

Unsupervised Machine Learning

● Challenges:
○ Evaluation: There are no clear evaluation metrics for unsupervised learning, making it
difficult to objectively compare different models or algorithms.
○ Interpretability: Understanding the meaning or interpretation of the patterns or clusters
discovered by unsupervised learning can be challenging, especially for complex models
or high-dimensional data.
○ Scalability: Some unsupervised learning algorithms can be computationally expensive
and difficult to scale to large datasets or high-dimensional data.
○ Determining the number of clusters: Determining the optimal number of clusters in a
clustering algorithm can be challenging and subjective.
○ Data preprocessing: Preprocessing and cleaning the data can be time-consuming and
require domain expertise, especially for complex or unstructured data like images or text.

15
 Machine Learning in Hindi

Practical Aspect: Tutorials

● Tutorials [Week 2 onwards] on Google Colab (Any other interface you are comfortable with,
is acceptable)
● Expected Basic Python. Should be comfortable with
○ Libraries: Numpy, Pandas, Matplotlib/Seaborn
○ Data types: Numpy Arrays, Pandas DataFrames, Lists, Dictionaries
○ Visualizations: Histogram, Scatter Plot, Box Plot, Line Plot, etc.
● Concepts:
○ Train-Test Split
○ Evaluation Metrics
○ Exploratory Data Analysis (EDA)
 Machine Learning in Hindi

Let us Try!
Question: Which of the following is an example of unsupervised machine learning?
A) Predicting stock prices
B) Recognizing handwritten digits
C) Clustering customer segments
D) Identifying spam emails

You may pause the video and try.

17
 Machine Learning in Hindi

Let us Try!
Question: Which of the following is an example of unsupervised machine learning?
A) Predicting stock prices
B) Recognizing handwritten digits
C) Clustering customer segments
D) Identifying spam emails

Explanation:
Predicting stock prices (A) and recognizing handwritten digits (B) are examples of supervised
machine learning tasks, where the algorithm learns to predict an output variable based on known
ground truth of the training samples. Identifying spam emails (D) is also a supervised machine
learning task. Clustering customer segments (C), on the other hand, is an example of unsupervised
machine learning, where the algorithm learns to group similar training data samples together based
on their features, without being explicitly told what the correct output groups should be.

18
 Machine Learning in Hindi

To Summarize
What we learnt in this module is as follows:

● Formal introduction to ML
● Supervised ML
● Unsupervised ML
● Practical Aspects
● Solved a question

19