PGP 1445 Mlai

Post Graduate Programme
in Artificial Intelligence &

Machine Learning
for Working
Professionals
INDEX
Programme introduction 01
Who should apply 02
The industry forecast 03
Programme highlights 04
Programme objectives 05
Mode of learning 06
Student learning outcomes 08
Programme structure 09
Programme curriculum
Course 1 Regression 10
Course 2 Feature Engineering 12

Course 3 Classification 14
Course 4 Unsupervised Learning & 17
Association Rule Mining
Course 5 Text Mining 20

Course 6 Deep Learning and ANN 22
Course 7 Capstone Project 24
Eligibility criteria 25
Fee structure 26
Mode of examination 27
How to apply 28
Program Introduction
With growing clutches of digitization and digital transformation initiatives

across several organisations in India, the demand for AI talent is expected
to skyrocket.
According to the Fortune Business Insights forecast, the global Machine

Learning market size was valued at $19.20 billion in 2022 and it is
expected to grow from $26.03 billion in 2023 to $225.91 billion by 2030.
So, get ready to make the most of it.
The 11-month Post Graduate Certificate Programme in Artificial Intelligence

and Machine Learning by BITS Pilani Work Integrated Learning
Programmes is designed to help working professionals like you develop a
deeper understanding of AI and ML and get equipped with knowledge on its
various building blocks.
01 Post Graduate Programme in Artificial Intelligence & Machine Learning APPLY NOW
Who Should Apply?
The programme is
designed for technology
professionals who wish
to advance their career
as a specialist in
Artificial Intelligence and
Machine Learning.
Professionals who wish

to transition to roles
such as Data Scientist,
Machine Learning
Engineer, AI Product
Manager, and Applied
ML Scientist should
consider applying to this
programme.
The Industry Forecast
● As per Forrester, over 50% of tech enterprises have already implemented or are
in the process of implementing AI & ML.
● By 2025, AI & Machine Learning will automate IT growth by 71%, as reported
by Forbes.
● The AI & ML market is expected to grow globally at a CAGR of 37.95% in
2020-26.
● AI & ML Digital skills to contribute as much as 25% in the overall hiring in the IT
sector.
Programme Highlights
11-month Post Graduate Access to BITS Pilani instructors

certificate programme for working through technology-enabled contact
professionals. classes which can be accessed
from anywhere, Q&A support, and
discussion forums.
Comprehensive and rigorous

Opportunity to become a
curriculum covering key
member of an elite & global
concepts and technologies of
community of BITS Pilani
Artificial Intelligence and
Alumni.
Machine Learning.
An 8-week Capstone project

where you will work towards
solving a Data Science Fee submission option using
related business problem easy- EMI with 0% interest and 0
under the mentorship of BITS down payment.
Pilani faculty members and
senior industry practitioners.
Two Immersion modules of 1/2

days each at a BITS Pilani
Campus or Online, during which
participants will visit the Campus to
interact with their peers and learn
together from BITS faculty.
Programme Objectives
Enable working professionals to Produce professionals with a

be industry-ready in the AI&ML strong algorithmic perspective
space. of AI&ML.
Provide a comprehensive Provide a deeper understanding

understanding of the data science of AI&ML techniques to enhance
pipeline. informed decision-making.
Provide hands-on to solve real-life

AI&ML problems.
Mode of Learning
The Mode of Learning used in this programme is called - Work Integrated Learning.
Internationally, Work Integrated Learning (WIL) is defined as "An educational approach
involving three parties - the student, educational institution, and employer organization(s)
- consisting of authentic work-focused experiences as an intentional component of the
curriculum. Students learn through active engagement in purposeful work tasks, which
enable the integration of theory with meaningful practice that is relevant to the students'
discipline of study and/or professional development*.
An education model can be considered as WIL if and only if:

1. The programs are designed and developed by the institute in collaboration with
industry.
2. Work-focused experiences form an active part of the curriculum.
3. The program structure, pedagogy and assessment enable integration of theory-with
relevant practice.
The innovative Work Integrated Learning Programs (WILP) of BITS Pilani are quite
aligned with the above definition and requirements. The programs are designed in
collaboration with its industry partners, subject matter experts from industry and
academia that enable the students to remain relevant in their chosen profession, grow in
their career and retain the habit of lifelong learning. Case studies, simulation exercises,
labs and projects further strengthen this integration.
The WILP of BITS Pilani is comparable to its campus-based programs in terms of

structure, rigor, instruction, labs, assessment, faculty profile and learning support. The
pervasive adoption of technology in all its academic processes makes the same
high-quality education of BITS Pilani available to the aspirants at scale with the required
flexibility.
Key Benefits of BITS Pilani WILP
1. Students can pursue the programme without any career break and along with
the job.
2. The programme curriculum is highly relevant to sectors, industries and organisations

they work for.
3. Effective use of technology to deliver a range of learning interventions such as

faculty contact sessions, asynchronous learning, remote, virtual and cloud labs,
learner support, peer to peer collaboration etc.
4. Contact sessions with faculty take place mostly over weekends or after business
hours and are conducted over a technology platform that can be accessed from
anywhere.
5. Comprehensive examinations will be scheduled at the end of Course 3 (for courses

1,2, 3) and Course 6 (for courses 4, 5,6) and are conducted mostly at designated
examination centres distributed across the country. Click Here.
6. Learners can access engaging learning material which includes recorded lectures
from BITS Pilani faculty members, course handouts and recorded lab content where
applicable.
Student
Learning Outcomes
After completing this programme, the
student will be able to Decide whether
AI&ML techniques are applicable to a
given business problem and articulate
their benefits thereof.
Collect data, apply

Identify appropriate
Formulate business pre-processing
techniques to solve the
problems as AI&ML techniques, and
formulated AI & ML
Problems. perform visual data
problem.
exploration.
Implement and evaluate

the relevant algorithms
Apply Supervised
using appropriate
Learning, Unsupervised
programming languages
learning, Text Mining and
and libraries. Interpret
Deep Learning
and present the
techniques.
model/analysis to the
stakeholders.
Programme Structure
The 11-month Post Graduate Programme in Artificial Intelligence and Machine

Learning consists of 6 Courses and a Capstone Project.
Course 1: Regression 5 weeks
Course 2: Feature Engineering 4 weeks
Course 3: Classification 9 weeks
Course 4: Unsupervised Learning & Association Rule Mining 7 weeks
Course 5: Text Mining 5 weeks
Course 6: Deep Learning and ANN 6 weeks
Course 7: Capstone Project 8 weeks
● In addition to the Curriculum above, participants will have the option of

taking an optional course on Python at the beginning of the Programme.
● This will allow participants to revisit essential concepts that will help in all
other courses during the programme.
● Topics covered include Introduction to Python programming and installation,
Data Types, Program constructs, Numpy, Pandas, Matplotlib, and
Debugging python programs.
Programme Curriculum
Course 1
Regression
● Regression is a type of supervised learning technique wherein the target attribute is a
continuous variable.
● This course focuses on developing a deeper understanding of regression models both
from theoretical and implementation perspectives.
● The model selection and performance measures will be discussed in this course.
● The issues with regression models like overfitting and the ways of combatting
overfitting like ridge and lasso regression will be illustrated in this course.
● The interpretability/explicability of the models will also be discussed.
The course aims to provide:

● Comprehensive algorithmic perspective of building regression models.
● Deeper understanding of overfitting and ways to combat overfitting.
● Competence to select appropriate model and performance measures.
● Hands-on to solve real life regression problems.
● Skill to interpret the predicted model.
Learning Outcomes
Upon completion of this course, participants will be able to:
1. Build appropriate regression model for a given real life business problem.
2. Demonstrate the capability to select suitable degree of the polynomial regression and
performance measures.
3. Suggest appropriate methods to combat overfitting.
4. Interpret the regression model.
Regression
Topics Covered
1. Introduction to six modules of the programme
Overview of 2. Programme Objectives & Learning outcomes
certificate programme 3. Evaluation of the courses (Quizzes/Assignments/Tests)
in ML & AI 4. ML&AI in today’s world
5. A real life ML&AI project and value of it to the business
1. Introduction to Supervised Learning

2. Introduction to Regression and Classification
Introduction to 3. Linear and Polynomial Regression
Regression 4. Error Function for Linear Regression
5. Introduction to Matrix Theory
6. Solving Simultaneous Equation with Matrices
1. Mathematical Foundations 1 – Maxima and minima of function of one variable

2. Mathematical Foundations 2 - rank, Eigen values and eigen vectors, positive &
negative definite/semi- definite matrices
3. Mathematical Foundations 3 - Maxima and minima of function of several
Building Simple Linear variables
Regression Models 4. Convexity of errors function
5. Building simple linear regression model by solving normal equations
6. Gradient descent algorithm
7. Gradient descent algorithm for a simplistic case
8. Gradient descent, stochastic & mini-batch gradient descent algorithms
1. Probability Foundations 1 – Discrete probability distributions

2. Probability Foundations 2 - Continuous probability distributions, normal
distribution and t distributions
Assessing Accuracy
3. Accuracy of the coefficient estimates of the simple linear regression models
of Simple Linear 4. Dependency of the dependent variable (target) variable on the independent
Regression Models variable (feature)
5. Accuracy of the simple regression model – RSE
6. Accuracy of the simple regression model – R^2
1. Building multiple linear regression model by solving normal equations

2. Building multiple linear regression model by gradient descent algorithms
Building Multiple
3. Performance measure for multiple linear regression models
Linear Regression
4. Feature selection algorithms for multiple linear regression models
Models 5. Forward and backward feature selection algorithms for multiple linear
regression models
1. Polynomial Curve Fitting – Finding suitable degree of the polynomial for the problem
Building Polynomial 2. Overfitting
Regression Models, 3. Ridge Regression and other ways of combatting overfitting
Overfitting and ways 4. Lasso Regression
to combat overfitting 5. Bias-Variance Decomposition 1
6. Bias-Variance Decomposition 2
Course 2
Feature Engineering
● Feature Engineering is an important step in developing and improving the
performance of Machine Learning models.
● This course covers Data wrangling techniques that help transform the raw data to an
appropriate form on which learning algorithms can be applied.
● The data preprocessing techniques like normalization, discretization, feature subset
selection etc. and dimension reduction techniques like PCA will be discussed.
● The different ways of visualizing the data like Box plots, Contour plots, Heat maps etc.
will be illustrated.
This course aims to help participants:

● Apply data wrangling for a given business problem.
● Identify and implement appropriate feature extraction and selection techniques.
● Apply data preprocessing techniques.
● Implement dimension reduction using PCA.
● Identify and implement appropriate visualization techniques.
Learning Outcomes
Upon completion of the course, participants will be able to:
1. Understand the importance of dimension reduction and apply appropriate techniques
given a real world application.
2. Compare the use of different similarity measures and Identify the appropriate
similarity measure to be used between 2 or more items, concepts, etc
3. Find and select appropriate visualization technique that answers a particular research
question.
Feature Engineering
Topics Covered
1. Introduction to Feature Engineering
Overview of Feature 2. Types of data and its sources
Engineering and 3. Data quality (Missing values, Noisy data)
Data Preprocessing 4. Aggregation and Sampling
5. Feature Creation
6. Data Transformation
1. Discretization
2. Supervised Discretization using Entropy
3. Binarization
4. Proximity measures for binary attributes
5. Proximity measures for Categorical attributes
Proximity measures 6. Proximity measures for Continuous attributes and Nonmetric measures
and Feature Subset 7. Curse of Dimensionality
Selection 8. Feature Subset Selection
9. Feature selection using Filter Methods
10. Feature selection using Chi Squared Test
11. Feature selection using Information Theoretic Measures
12. Feature selection using Fisher Score
13. Feature selection using wrapper methods
1. Statistics foundations
Dimension 2. Introduction to Dimension reduction
Reduction 3. Formulation of Principal Component Analysis (PCA) using Maximum Variance
4. PCA mathematical derivation
1. Data Visualization and Statistical Analysis

2. Charts and Plots
Visualization 3. Parallel Coordinates
4. t-SNE Plot
5. Data Visualization Use case and Feature Engineering
Course 3
Classification
● Classification is a type of supervised learning technique wherein the target attribute
takes discrete values.
● This course emphasizes the three types of techniques to solve classification problems
– discriminant function, generative and probabilistic discriminative approaches.
● This course lays down a strong foundation on the algorithmic perspective of popular
classification algorithms - k-NN, Na ve Bayes, Decision Tree, Logistic Regression and
SVM.
● The implementation details of these models along with tuning of parameters will be
illustrated. The ensemble methods, bagging, boosting, Random Forest and eXtreme
Gradient Boosting will be taught.
● The interpretability/explicability of the models will also be discussed.
The course aims to:

1. A deeper understanding of three types of techniques to solve classification problems.
2. Comprehensive algorithmic perspective of popular classification algorithms.
3. Hands-on to solve real-life classification problems.
4. The skill to interpret the predicted model.
5. The competence to build ensemble classifiers using well-known techniques.
Learning Outcomes
1. Build appropriate classifier for a given real life business problem
2. Demonstrate the capability to understand classification algorithms deeply and fine
tuning the parameters therein to enhance performance of the classifier
3. Build ensemble classifier using well known techniques
4. Interpret the regression model
Classification
Topics Covered
1. Introduction to Classification
2. Types of classification algorithms - Discriminant Functions, Probabilistic
Overview of the Generative models and Probabilistic Discriminative models, Tree based models
Classification Module 3. Classification Algorithms covered in the course and type of these algorithms
4. Applications of classification and case study of the course
1. kNN Classifier
Nearest-neighbour 2. Measures of prediction accuracies of classifiers – precision, recall, AUC of ROC etc.
Methods 3. Finding optimal k
4. Python Implementation of kNN
1. Probability Foundations – Discrete & Continuous Random Variables, Conditional

Independence, Bayes Theorem (1)
2. Probability Foundations – Discrete & Continuous Random Variables, Conditional
Independence, Bayes Theorem (2)
Naïve Bayes 3. Naïve Bayes Classifier – Derivation
Classifier 4. An illustrative example
5. Python implementation of Na ve Bayes Classifier
6. Na ve Bayes Classifier is a generative model
7. Advantages of Naïve Bayes Classifier and when to use Na ve Bayes Classifier?
8. Interpretability of Na ve Bayes Classifier
1. Significance of Sigmoid function and finding its derivative

2. Statistics Foundations – Maximum likelihood estimation
3. Cross entropy error function for logistic regression and its optimal solution
Logistic 4. Logistic Regression is probabilistic discriminative model and an illustrative example
Regression 5. Implementation of logistic Regression using Python
6. Decision boundary of logistic regression
7. Overfitting of logistic regression and counter measures
8. Interpretability of logistic regression
1. Decision Tree Representation

2. Entropy and Information Gain for an attribute
3. Search in Hypothesis space, ID3 Algorithm for decision tree learning
4. Implementation of Decision Tree using Python
Decision Tree 5. Prefer short hypothesis to longer ones, Occam’s razor
6. Overfitting in Decision Tree
7. Reduced Error Pruning and Rule post pruning
8. Alternative measures for selecting attributes
9. Interpretability of Decision Tree
Classification
Topics Covered
1. Constrained and Unconstrained Optimization
Optimization
2. Primal and Dual of an optimization problem
Foundations for
3. Quadratic Programming
Support Vector 4. KKT conditions
Machines 5. Lagrange Multiplier
1. Understanding the spirit and significance of maximum margin classifier

2. Posing an optimization problem for SVM in non-overlapping class scenario
3. Converting the constrained optimization problem into unconstrained using
Support Vector Legrange multipliers
Machines 4. Dual of the optimization problem
5. Appreciation of sparse kernel machine and support vectors in the solution of the
optimization problem
6. Implementation of SVM in python
1. Issues of overlapping class distribution for SVM

Support Vector
2. Posing an optimization problem for SVM in overlapping class scenario
Machines in
3. Solving the optimization problem using Legrange multipliers, dual representations
overlapping class 4. Kernel Trick and Mercer’s theorem
distributions & 5. Techniques for constructing Kernels and advantages of Kernels in SVM
Kernels 6. Implementation of SVM using different kernels
1. Rational for Ensemble Method

2. Methods for constructing an Ensemble Classifier
3. Bagging, Boosting, AdaBoost
Ensemble Methods 4. Random Forest
5. eXtreme Gradient Boosting (XGBoost)
6. Python Implementation of Random Forest and XGBoost
7. Class Imbalance Problem & approaches to solve it
Course 4
Unsupervised Learning & Association Rule Mining
● Unsupervised learning algorithms find regularities of the given dataset in the absence
of explicit labels or supervised outputs for the data points.
● Clustering is an unsupervised learning task whose objective is to find natural grouping
present in the data.
● This course covers various clustering algorithms like K-Means, EM Algorithm, Single
Linkage Algorithm, Complete Linkage algorithm and DBSCAN.
● Various ways of assessing the quality of clustering and detecting outliers are
discussed.
● The typical industrial applications of unsupervised learning algorithms are covered as
well.
● HMM is introduced in the context of performing time series prediction and the role of
EM algorithm in estimating the parameters are discussed.
● The other part of the course introduces an important class of algorithms to learn
association or discover dependencies between the data items, known as learning
association rules.
● We discuss apriori algorithm and different metrics to measure the interestingness of
the rules.
The course aims to introduce:

1. Unsupervised learning, various unsupervised learning algorithms
2. Association rule learning and apriori algorithms
3. Time series data and use HMM to solve various tasks involving Time series data
Learning Outcomes
1. Understand various algorithms for clustering, association rule mining and the role of
HMM in time series prediction tasks.
2. Analyse the problem and provide learning solutions using the algorithms covered in
this course.
3. Apply the learning algorithms suitably to solve various tasks including anomaly
detection, parameter estimation, segmentation etc.
4. Analyse the given problem, decide the suitability of association rule learning
technique to solve this and provide a solution.
5. Apply the HMM suitably to solve problems involving time series data.
Topics Covered
a. Unsupervised Learning - Introduction - Applications- Clustering as an
Introduction to unsupervised learning task - Defining clustering
Unsupervised b. Introducing Various ways to solve clustering problem ( similarity based, density
Learning, Clustering based, hierarchical, graph theoretic based) - Notion of quality of clustering
c. Overview of clustering algorithms to be covered in this course
Introducing the
clustering case study to a. overview of the data set to be used
be used throughout the b. Exploring this data using Python
course for assignments
a. K-Means Algorithm
b. Discussion on Various Initializations, Standardizing Attributes (for eg- z-score)
K-Means Algorithm & Convergence
c. Python Implementation
d. Applications of using K-means with Images, videos, documents
a. Online stochastic version of k-means (with sequential update) - Discussions on quality

K-Means - Variations of clustering / convergence - Applications
b. Mini-Batch K-Means - Discussions on quality of clustering / convergence - Applications
a. Outliers and Clustering - Overview.

Detecting Outliers b. Using K-means to detect outliers
c. Python Implementations
a. Mixtures of Gaussians (MoG) - Applications, modelled as MoG

b. Using Maximum Likelihood to estimate mixture densities - Issues
EM Algorithm c. EM Algorithm for Gaussian mixtures
i. Derivation
ii. Illustration ( using a problem involving mixture of two gaussians)
+ Python Implementations
iii. Applications
d. Relationship to K-Means Algorithm
Clustering for
Customer
Segmentation -
[ Pre-Recorded
Industry Talk ]
a. Introduction to hierarchical clustering

b. Agglomerative Clustering Vs Divisive Clustering
Hierarchical c. Distance Measures (Minimum distance, Maximum Distance, Mean
Clustering Distance, Average Distance)
d. Algorithms
i. Single linkage, Complete Linkage algorithm
ii. Demonstration in python
iii. Discussion on Termination, efficiency, applications
Topics Covered
Density Based a. Density based approach to clustering - Introduction
Clustering b. DBSCAN - Density, Density-reachability, Density-connectivity
c. DBSCAN Algorithm
d. Performance & scalability
e. Demonstration using Python
Clustering for
Anomaly Detection -
[ Pre-Recorded
Industry Talk ]
Assessing Quality a. Cluster Validity Evaluation (measuring compactness, separation, cluster overlap, etc)
of Clustering b. Stability of Results from clustering algorithms
c. Determining number of clusters
Significance of
Clustering - Interpreting/
summarizing Clusters
by businesses - [Pre-
Recorded Industry Talk]
a. Market Basket Analysis - Use cases

Association Rule b. Terminologies / Measures - association rules, support, confidence ,k-itemset,
Mining Frequent itemsets, closed item sets
c. Discussion on computational complexity in generating the itemsets
a. Algorithm
b. Generating Association Rules from frequent itemsets
Apriori Algorithm c. Efficiency Issues and few ways to address it.
d. Evaluating interestingness of patterns
e. Demonstration of Apriori algorithm using python for a practical use case
a. Introduction
i. Introduction to time series data
Time series Prediction
ii. Time Series prediction applications (eg predicting stock prices, fraud
and Markov Process
detection, applications in text and speech processing)
b. (discrete) Markov Processes - Overview and Terminologies
a. i. Introduction
ii. Evaluation Problem - Given a model, evaluate the probability of observing the
sequence - (forward-backward Procedure)
Hidden Markov Model iii. Finding most likely state sequence explaining time series data - Viterbi Algorithm
iv. How to learn Model parameters - An application of EM Algorithm
b. Case Study: Introduce a problem from an application domain- solution using HMM -
Python Implementation
Course 5
Text Mining
● In today’s world, 80% of the data generated by enterprises is unstructured or
semi-structured in the form of Emails, Surveys, Feedback etc. wherein most of the
data is in the form of text.
● This course aims to equip students with adequate knowledge in extracting the
relevant text data and skills to identify patterns therein.
● This course covers topics like converting documents to vectors using TF-IDF, Parts of
Speech Tagging, Topic modelling using LDA, sentiment analysis and recommender
systems.
This course is designed to help participants:

1. Convert documents into vectors using TF-IDF and compute similarities.
2. Implement topics modelling using LDA.
3. Apply sentiment analysis.
4. Implement recommender systems.
Learning Outcomes
Upon completion of the course, participants will gain:
1. Knowledge of basic pipeline of Information retrieval and vectorising the documents
using TD-IDF and its implementation.
2. A mathematical foundation on modelling Parts of speech (POS) tagging using Hidden
Markov Model and hands on experience of implementing POS tagging.
3. A mathematical foundation on modelling Latent Dirichlet Analysis (LDA) for topic
modelling and hands on experience of implementing it.
4. Knowledge on sentiment analysis and its applications.
5. A mathematical foundation of Recommender Systems and hands on experience of
implementing it.
Text Mining
Topics Covered
1. Introduction to Text Mining
2. Binary term incidence matrix
Information Retrieval 3. Information Retrieval Pipeline
4. Inverted Index Construction
5. Merge Algorithm and Query Optimization
6. Tolerant Retrieval using Normalization, Query expansion, Stemming,
Lemmatization, Wild card query using K-Gram index
7. Ranked Retrieval using TF-IDF and Cosine score
1. Introduction to Part of speech tagging

Part of Speech
2. Part of speech tagging using HMM-1
Tagging
3. Implementing POS Tagging in Python
1. Mathematical foundations for LDA : Multinomial and Dirichlet distributions-1

2. Mathematical foundations for LDA : Multinomial and Dirichlet distributions-2
3. Intuition behind LDA
Topic modelling
4. LDA Generative model
using LDA
5. Probabilistic Graphical Models
6. Latent Dirichlet Allocation
7. Implementing LDA in Python
1. Sentiment Analysis
2. Subjectivity Analysis
Introduction to 3. Topic Extraction
Sentiment 4. Product Reviews
Analysis 5. Opinion Retrieval and Spam
6. Opinion Summarization
7. Implementing Sentiment Analysis in Python
1. Introduction to Recommender Systems

2. Collaborative filtering
2.1 User based Collaborative filtering
2.2 Item based Collaborative filtering
Recommender 2.3 Matrix factorization using Singular Value Decomposition
Systems 2.4 Latent Factor Models
3. Metrics used for evaluating Recommender Systems
4. Implementing Recommender System in Python
5. Industry talk on application of Recommender Systems
Course 6
Deep Learning and ANN
● Due to the availability of low-cost hardware in the last decade, several involved neural
network approaches had been explored to advance the performance of many
state-of-the-art machine learning problems such as image searching, understanding,
medical applications, autonomous vehicles such as drones and self-driving cars etc.
● In this course students will be exposed to the details of neural networks as well as
deep learning architectures.
● This course gives an algorithmic perspective and implementation details of ANN,

RNN, LSTM, CNN, RCNN, Faster RCNN, Autoencoders, Generative deep learning
models like VAE and GAN etc.
This course is designed to help participants:

1. Learn ML approaches that are inspired by the human brain.
2. Design shallow and deep networks using non-linear activation functions.
3. Understand sequence-aware neural networks for text/voice/video etc.
4. Learn Convolution Neural Networks to solve typical computer vision problems.
5. Learn how data compression could be achieved using a network of neurons.
Learning Outcomes
1. Design a multilayer ANN for an appropriate learning problem.
2. Design multilayer neural network architecture for text/voice/video or any sequence
data.
3. Appreciate the concept of end-to-end learning and feature abstraction
4. Apply convolution and highlight useful parts in a scene for computer vision
applications.
5. Apply generative approach to solve ML problems.
Deep Learning and ANN
Topics Covered
1. Introduction and Background
2. Discrimination power of single neuron
Artificial Neural 3. Training a single perceptron (delta rule)
Network 4. Multilayer Neural Networks
5. Activation functions and Loss functions
6. Backpropagation
1. Introduction to end to end learning

Deep Learning 2. Abstractions of features using deep layers
3. Hyper parameter tuning
4. Regularization for Deep Learning
5. Dropout
1. CNN
2. Pooling
Convolution 3. Variants of pooling functions
Networks with
4. CNN with Fully connected Networks
Deep Learning
5. RCNN
6. Faster RCNN
1. Architecture of RNN
2. Unfolding of RNN
Sequence Modeling in 3. Training RNN
Neural Network 4. LSTM
5. LSTM and its applications
1. Undercomplete Autoencoders
Autoencoders with 2. Regularized Autoencoders
Deep Learning 3. Variational autoencoders
4. Applications of Autoencoders
Generative deep 1. GAN

learning models 2. Applications of GAN
Course 7
Capstone Project
● During the 8-week Capstone Project, participants will work in teams to design and
solve a real-world business problem encompassing data science pipeline using
AI&ML techniques.
● Participants will be required to identify applicable AI and ML solutions and apply these
solutions to arrive at outcomes.
● Through each phase of the project, participants will be mentored by BITS Pilani
faculty members and senior Industry practitioners using a rigourous and structured
framework and will receive regular feedback on their progress.
Eligibility Criteria
● Employed professionals holding BE/ B.Tech. or equivalent, one year's work
experience and working in relevant fields are eligible to apply.
● Candidates holding M.Sc. in Mathematics or Statistics, one year's work experience

and working in relevant roles are also eligible to apply to this programme.
● A working knowledge of languages such as Python is recommended.
● Applicants without sufficient exposure to Python programming language will have to

complete a refresher course in Python as a part of the programme before the start
of Course 1.
Fee Structure
Fee Structure for students admitted in Academic Year 2024-2025 is as follows:
Block Amount Programme Fee

(one time) (remainder )
`25,000 `2,20,000
(within 7 days of (within 15 days of

receipt of provisional receipt of Final
Admission Offer Letter) Admission Offer Letter)
Total Programme Fee is INR 2,45,000 (including GST)

Easy Monthly Payment Option with 0% Interest and 0 Down Payment
Instant EMI option with 0% interest and 0 Down Payment is now available
that allows you to pay programme fee in an easy and convenient way.
● Instant online approval in seconds

● No Credit Cards/ CIBIL score required
● Easy & Secure online process using Aadhaar and PAN number
● Anyone with a Salary Account with Netbanking can apply
Admission Open.
Click here to learn more
Last date to apply is Sept 16, 2024
All the above fees are non-refundable.

Important: For every course in the programme, institute will recommend textbooks,
students would need to procure these textbooks on their own.
Mode of Examination
Examinations Mode Options for Post Graduate Programme in Artificial Intelligence &
Machine Learning applicable for students admitted in Batch starting from Oct / Nov 2024.
Comprehensive examinations will

be conducted for each Course
Mode of Examinations in the programme
applicable for students These exams are typically
admitted in conducted at the end of Course 3
(for Courses 1, 2, 3) and for
Batch starting in Courses 4, 5, 6 before starting the
Oct / Nov 2024. Capstone Project. These
examinations are mostly scheduled
on Friday, Saturday or Sunday.
Students need to appear in In addition to the

person for taking the comprehensive examinations
examinations at the for each course, there will
institution’s designated also be Quizzes/Assignments
examination centres as per conducted online on the
the examination schedule, Learning Management
Instructions, rules and System (LMS) as per the
guidelines announced before course plan in which the
every examination. students need to participate.
Students can take their examination at any of our 33 designated examination centres in India at the
following locations:
● South Zone: Bangalore - North, Bangalore - Central, Bangalore - South, Bangalore - East, Chennai -
North, Chennai - Central , Chennai - South, Hyderabad, Secunderabad, Vijayawada, Visakhapatnam,
Kochi, Thiruvananthapuram and Coimbatore.
● North Zone: Delhi, Gurugram, Noida, Jaipur, Chandigarh, Lucknow and Pilani.
● West Zone: Mumbai, Navi-Mumbai, Pune, Pune - Pimpri Chinchwad, Goa, Ahmedabad, Indore and
Nagpur.
● East Zone: Kolkata, Bhubaneswar, Guwahati and Jamshedpur.
In addition to these locations, the Institution also has a designated examination
centre in Dubai.
How to Apply
Click here to Create your login at the

You will receive a
apply now Application Center by Provisional
entering your unique Admission Offer
Email id and create a Letter within 2 days
password of your of receipt of your
choice. Application Form.
Upon receiving the Provisional Admission Offer Letter, you will need to submit the
following within 7 days using the Online Application Center:
� Block amount of INR 25,000
� Scanned copy of Passport size photograph.
� Scanned copy of self-attested Graduation degree certificate and marksheets.
� Proof of ID (Govt. issued ID such as Driving License, Passport, Aadhaar, Voter ID, etc.)
� Proof of employment, such as Work Experience Certification from current employer.
Upon receiving the Block Amount and other supporting documents, you will receive a Final
Admission Offer Letter. You will need to submit the First Installment (INR 2,20,000) within 15 days
of receipt of this letter. For details on No-cost EMI option with 0% interest, Click Here.
Upon receipt of the remaining First Installment, you will receive your BITS Student ID, detailed
programme schedule and access to the learning.
B2C_25_07_2024
Let's start a conversation
to ignite the change you desire
https://bits-pilani-wilp.ac.in
Call: 080-48767777
admission@wilp.bits-pilani.ac.in

PGP 1445 Mlai

Uploaded by

Copyright:

Available Formats

PGP 1445 Mlai

Uploaded by

Document Information

Original Title

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

PGP 1445 Mlai

Uploaded by

Copyright:

Available Formats

Post Graduate Programme

in Artificial Intelligence &

Course 2 Feature Engineering 12

Course 5 Text Mining 20

Course 7 Capstone Project 24

With growing clutches of digitization and digital transformation initiatives

According to the Fortune Business Insights forecast, the global Machine

The 11-month Post Graduate Certificate Programme in Artificial Intelligence

Professionals who wish

11-month Post Graduate Access to BITS Pilani instructors

Comprehensive and rigorous

An 8-week Capstone project

Two Immersion modules of 1/2

Enable working professionals to Produce professionals with a

Provide a comprehensive Provide a deeper understanding

Provide hands-on to solve real-life

An education model can be considered as WIL if and only if:

The WILP of BITS Pilani is comparable to its campus-based programs in terms of

2. The programme curriculum is highly relevant to sectors, industries and organisations

3. Effective use of technology to deliver a range of learning interventions such as

5. Comprehensive examinations will be scheduled at the end of Course 3 (for courses

Collect data, apply

Implement and evaluate

The 11-month Post Graduate Programme in Artificial Intelligence and Machine

Course 1: Regression 5 weeks

Course 2: Feature Engineering 4 weeks

Course 3: Classification 9 weeks

Course 4: Unsupervised Learning & Association Rule Mining 7 weeks

Course 5: Text Mining 5 weeks

Course 6: Deep Learning and ANN 6 weeks

Course 7: Capstone Project 8 weeks

● In addition to the Curriculum above, participants will have the option of

The course aims to provide:

1. Introduction to Supervised Learning

1. Mathematical Foundations 1 – Maxima and minima of function of one variable

1. Probability Foundations 1 – Discrete probability distributions

1. Building multiple linear regression model by solving normal equations

This course aims to help participants:

1. Data Visualization and Statistical Analysis

The course aims to:

1. Probability Foundations – Discrete & Continuous Random Variables, Conditional

1. Significance of Sigmoid function and finding its derivative

1. Decision Tree Representation

1. Understanding the spirit and significance of maximum margin classifier

1. Issues of overlapping class distribution for SVM

1. Rational for Ensemble Method

The course aims to introduce:

a. Online stochastic version of k-means (with sequential update) - Discussions on quality

a. Outliers and Clustering - Overview.

a. Mixtures of Gaussians (MoG) - Applications, modelled as MoG

a. Introduction to hierarchical clustering

a. Market Basket Analysis - Use cases

This course is designed to help participants:

1. Introduction to Part of speech tagging

1. Mathematical foundations for LDA : Multinomial and Dirichlet distributions-1

1. Introduction to Recommender Systems

● This course gives an algorithmic perspective and implementation details of ANN,

This course is designed to help participants:

1. Introduction to end to end learning