Interested in this College?
Get updates on Eligibility, Admission, Placements Fees Structure

Quick Facts

Medium Of Instructions	Mode Of Learning	Mode Of Delivery
English	Self Study, Virtual Classroom	Video and Text Based

Course Overview

The Post Graduate Programme in Artificial Intelligence and Machine Learning programme is specifically designed to make AI & ML knowledge accessible to professionals without taking a career break. The course ensures implementable AI & ML techniques to the candidates. The training is structured towards shaping the candidates into industry-ready AI & ML professionals. The Post Graduate Programme in Artificial Intelligence and Machine Learning online course has two modes: theoretical explanation and practical guidance.

The Post Graduate Programme in Artificial Intelligence and Machine Learning certification is an 11-month online programme. It consists of 6 courses and a capstone project. It will also provide an additional course on Python in the beginning. This course offers live interaction with the Post Graduate Programme in Artificial Intelligence and Machine Learning certification course instructors given in the Post Graduate Programme in Artificial Intelligence and Machine Learning programme schedules. There will be two campus immersion modules for two days in the Hyderabad/ Goa campus of BITS Pilani where the participants can go and interact with their peers and also learn from BITS faculty.

The Post Graduate Programme in Artificial Intelligence and Machine Learning by Birla Institute of Technology, Pilani certification is given on successful completion of the 11-month programme.

The Highlights

Digital learning through videos
EMI option with 0% interest available
Offered by BITS Pilani
Campus Immersion in Hyderabad/Goa
11-month programme
Periodic online interaction with instructors
Online and offline exam modes are available
UGC approved course
Capstone project for 6 weeks

Programme Offerings

Live online lectures
discussion forum
assignments
Periodic live interactions
Campus immersion
Capstone Project

Courses and Certificate Fees

Certificate Availability	Certificate Providing Authority
yes	BITS Pilani

The candidates applying for this course are requested to check the course fee details in these points given below-

The candidates will have to pay Post Graduate Programme in Artificial Intelligence and Machine Learning fees of Rs. 2,45,000 with GST in total for the entire course.
The candidates will have to pay the block amount of Rs. 25,000 within 7 days of receiving the provisional admission offer letter.
The remaining fee of Rs. 2,20,000 will have to be paid within 15 days of receiving the final admission offer letter.
This programme also offers a No-cost EMI option.

Post Graduate Programme in Artificial Intelligence and Machine Learning fees

Fee Category	Amount in INR
Block amount	Rs. 25,000
Registration amount	Rs. 2,20,000

Eligibility Criteria

Work experience

Candidates with work experience in technology or science-related fields are eligible to apply for this course. Candidates with a working knowledge of Python are recommended.

Education

The candidates applying for the training are expected to have B.Tech or B.E., M.Sc in Statistics or Mathematics. Apart from these educational qualifications, working knowledge of Python is recommended.

Certification qualifying details

The Post Graduate Programme in Artificial Intelligence and Machine Learning certification will be provided to the candidates on successful completion of the programme which includes 6 courses and a Capstone project. The candidates will also have to make sure they have completely paid the full Post Graduate Programme in Artificial Intelligence and Machine Learning fees.

What you will learn

Machine learningKnowledge of PythonKnowledge of Artificial Intelligence

The candidates interested in the course will be benefited by the Post Graduate Programme in Artificial Intelligence and Machine Learning certification syllabus and practical guidance. These are the following learning outcomes for this course-

The candidates will learn how to formulate business-related problems as AI & ML problems.
The pre-processing techniques will be shown for data to the candidates.
The ways to collect data will be taught under the Post Graduate Programme in Artificial Intelligence and Machine Learning training.
The candidates will learn to use Python and implement and compare the candidates.
Interpreting and presenting a model.
Live interactions with data scientists will help the candidates to know more about the existing technologies.
The application of Supervised learning, Unsupervised learning, Deep learning and Visualization techniques will be a part of the training.
The candidates will get ample opportunities to revise their course learning through its assignments, discussion forum and a capstone project as a part of the Post Graduate Programme in Artificial Intelligence and Machine Learning online course.
The classification of algorithms and their comprehension will be emphasised.
ANN, Text Mining introduction and application can also be found.

Who it is for

This programme is fitted best for these types of candidates given below-

The candidates who are working in the technology sector.
The candidates who have a working knowledge of python.
The candidates who are generally interested in AI & ML.

Admission Details

The interested candidates can apply for the course in these following steps-

Step 1: The candidates will have to open this URL- https://bits-pilani-wilp.ac.in/certification-programmes/pgp-ai-ml.php#how-to-apply

Step 2: The candidates will have to login by clicking on the resources option on the course page.

Step 3: For login, the candidates will have to give their email id.

Step 4: The candidates will receive a provisional admission offer letter in two days of receiving the application form.

Step 5: On receiving the provisional admission offer letter, the candidates will have to submit the documents and amount in 7 days through an online application centre.

Step 6: The next step is the final admission offer letter. The first instalment will have to be submitted within 15 days of receiving the letter. It can also be paid through 0% interest EMIs.

Step 7: After the instalment, the candidates will receive BITS student ID, the programme schedule and from there the candidates can start their learning process.

The Syllabus

Overview of certificate programme in ML & AI

Introduction to six modules of the programme
ML & AI in today’s world
A real-life ML & AI project and value of it to the business
Evaluation of the courses(quizzes/assignments/tests)
Programme objectives and learning outcomes

Introduction to regression

Linear polynomial regression
Regression vs Classification
Introduction to supervised learning
Overview of model building for linear regression
Applications and case study for the module

Mathematics foundation

Determinant & inverse of matrices, solving simultaneous equation
Convex function, necessary and sufficient condition for convexity of functions
Maxima and minima of univariate and multivariate functions
First and second derivatives of multivariate functions

Model building using least squares

Implementation in python (gradient & stochastic gradient descent methods)
Optimizing cost/ loss function stochastic gradient descent and batch gradient descent
Optimizing cost /loss function by gradient descent (I)
Optimizing cost/ loss function by gradient descent (II)
Implementation in python
Optimizing cost/loss function by solving normal equations
Convexity of the cost/loss function
Cost/loss function for linear regression

Model accuracy and selection

Implementation in python
Polynomial regression- selecting the appropriate degree of the polynomial
Training data, testing data and cross-validation data
Bias-variance decomposition
Implementation in python
Measuring the quality of fit

Overfitting

Compare ridge vs lasso vs model without regularization with a case study
Implementation in python (lasso)
Counters to control overfitting- lasso regression
Implementation in python (ridge)
Counters to control overfitting – ridge regression
Reasons to overfitting
Introduction to overfitting

Interpretability of regression models

Discussion on regression for a real-life business scenario
Interpretability of the regression built for the case study
Interpretability of the regression model through coefficients of the model
Statistics foundations- inferential statistics and hypothesis testing, significance tests p-values1
Statistics foundations- inferential statistics and hypothesis testing, significance tests p-values2

Overview of feature engineering

Implementation of scraper using python
Data quality (missing values, noisy data)
Types of data and its resources
Introduction to feature engineering

Data preprocessing

Similarities between attributes-1
Similarities between attributes-2
Implementing feature selection using python
Feature selection using wrapper methods
Feature selection using filter methods
Feature subset selection
Data transformation
Discretization and binarization
Feature creation
Aggregation and sampling

Dimensionality reduction

Industry talk on feature engineering for a problem domain
Implementing PCA using python
Principal component analysis (PCA) using minimum variance formulation 1
Principal component analysis (PCA) using minimum variance formulation 2
Principal component analysis (PCA) using minimum variance formulation 3
Introduction to dimension reduction
Statistics foundations (Variance, Covariance)

Visualization (industry expert)

TSNE
Parallel coordinates
Heat maps
Contour plots
Box/scatter plots
Bar charts/ pie charts
Histograms
Summary statistics
Industry talk on visualization

Overview of the classification module

Applications of classification and case study
Classification algorithms covered in the course and type of these algorithms
Types of classification algorithms- discriminant functions, probabilistic generative models and probabilistic discriminative models, tree based models
Introduction to classification

Nearest-neighbour methods

Python implementation of KNN
Finding optimal k
Measures of prediction accuracies of classifiers- precision, recall, AUC of ROC etc
kNN classifier

Naive bayes classifier

Interpretability of naive bayes classifier
Advantages of naive bayes classifier and when to use naive bayes classifier
Naïve bayes classifier is a generative model
Python implementation of naïve bayes classifier
Naïve bayes classifier – derivation
Probability foundations- discrete and continuous random variables, conditional independence, bayes theorem (1)
Probability foundations – discrete and continuous random variables, conditional independence, bayes theorem (2)

Logistic regression

Interpretability of logistic regression
Overfitting of logistic regression and countermeasures
Decision boundary of logistic regression
Implementation of logistic regression using python
Logistic regression is probabilistic discriminative model
Cross entropy error function for logistic regression and its optimal solution
Statistics foundation- maximum likelihood estimation
Significance of sigmoid function and finding its derivative

Decision tree

Interpretability of decision tree
Alternative measures for selecting attributes
Reduced error pruning and rule post pruning
Overfitting in the decision tree
Prefer short hypothesis to longer ones, occam’s razor
Implementation of logistic regression using python
Search in hypothesis space, ID3 algorithm for decision tree learning
Entropy and information gain for an attribute
Decision tree representation

Optimization foundations for support vector machines

Lagrange multiplier
KKT conditions
Quadratic programming
Primal and dual of an optimization problem
Constrained and unconstrained optimization

Support vector machines

Implementation of SVM in python
Appreciation of sparse kernel machine and support vectors in the solution of the optimization problem
Dual of the optimization problem
Converting the constrained optimization problem into unconstrained using Legrange multipliers
Posing an optimization problem for SVM in non-overlapping class scenario
Understanding the spirit and significance of maximum margin classifier

Support vector machines in overlapping class distributions and kernels

Implementation of SVM using different kernels
Techniques for constructing kernels and advantages of kernels in SVM
Kernel trick and mercer’s theorem
Solving the optimization problem using lagrange multipliers, dual representations
Posing an optimization problem for SVM in overlapping class scenario
Issues for overlapping class distribution for SVM

Ensemble methods

Class imbalance problem and approaches to solve it
Python implementation of random forest and XGBoost
eXtreme gradient Boosting(XGBoost)
Random forest
Bagging, boosting, adaboost
Methods for constructing an Ensemble classifier
Rational for ensemble method

Introduction to unsupervised learning, Clustering

Overview of clustering algorithms
Introducing various ways to solve clustering problem- notion of quality clustering
Unsupervised learning- introduction- applications- clustering as a unsupervised learning task-defining clustering

Case study

Introducing the clustering case study (to be identified) to be used throughout the course assignments- exploring this data using python, overview of the data set to be used.

K means algorithm

Demonstration in python
Applications of using k-means with images, videos, documents
Discussion on various initializations, standardizing attributes (for eg- z score) & convergence
K means algorithm

K means variation

Mini-batch k means- discussions on quality clustering/ convergence- applications
Online stochastic version of k means(with sequential update)- discussions on quality of clustering/ convergence-applications

Detecting outliers

Using k-means to detect outliers
Demonstration in python
Outliers and clustering- overview

Math fundamentals for EM algorithms

KL divergence
Jensen’s inequality

EM algorithm

Relationship to the k-means algorithm
EM algorithm for Gaussian mixtures- Derivation, Illustration of a problem using a mixture of two Gaussians and python, General form of EM algorithm and applications
Using maximum likelihood to estimate mixture densities- issues
Mixtures of Gaussians (MoG)- applications modelled as MoG

Hierarchical Clustering

Algorithms- Single linkage, complete linkage algorithm, Demonstrationin python, Discussion on termination, efficiency, applications
Distance measures
Agglomerative clustering vs Divisive clustering
Introduction to hierarchical clustering

Density based clustering

Demonstration using python
Performance & scalability
DBSCAN algorithm
DBSCAN- density, density- reachability, density- connectivity
Density based approach to clustering- introduction

Assessing Quality of clustering

Determining number of clusters
Stability of results from clustering algorithms
Cluster validity evaluation

Association rule mining

Discussion on computational complexity in generating the itemsets
Terminologies/ measures- association rules, support, confidence, k-itemset, frequent itemsets, closed itemsets
Market basket- analysis-use cases

Apriori algorithm

Demonstration of apriori algorithm using python for a practical use cases
Evaluating interestingness of patterns
Efficiency issues and few ways to address it
Generating association rules from frequent itemsets
Algorithm

Time series prediction and Markov process

Introduction- introduction to time series data, time series prediction applications
(discrete) Markov processes- overview terminologies

Hidden markov model

Learning model parameters- an application of EM algorithm
Finding most likely state sequence explaining time series data- viterbi algorithm
Evaluation problem- given a model, evaluate the probability of observing the sequence (forward-backward procedure)
Case study: Introduce a problem from an application domain- solution using HMM- python implementation/ demonstration

Document vectorization and parts of speech tagging

Ranked retrieval using TF-IDF and cosine score
Tolerant retrieval using normalization, query expansion, stemming, lemmatization, wild card query using K-gram index
Merge algorithm and query optimization
Inverted index construction
Information retrieval pipeline
Binary term incidence matrix
Introduction to text mining
Implementing POS tagging in python
Part of speech tagging using HMM-1
Introduction of part of speech tagging

Topic modelling using LDA

Implementing LDA in python
Latent Dirichlet allocation
Probabilistic graphical models
LDA generative model
Intuition behind LDA
Mathematical foundations for LDA- multinomial and dirichlet distributions 1
Mathematical foundations for LDA- multinomial and dirichlet distributions 2

Introduction to sentiment analysis

Implementing sentiment analysis in python
Opinion summarization
Opinion retrieval and spam
Product reviews
Topic extraction
Subjectivity analysis
Sentiment analysis

Recommender Systems

Industry talk on application of recommender systems
Implementing recommender system in python
Metrics used for evaluating recommender systems
Collaborative filtering- User based collaborative filtering, item based collaborative filtering, matrix factorization using singular value decomposition
Latent factor models

Artificial neural network

Backpropagation-1
Backpropagation-2
Activation functions and loss functions
Multiplayer neural networks
Training a single perceptron(delta rule)
Discrimination power of single neuron
Introduction and background

Sequence modeling in neural network

LSTM (1)
LSTM (2) and its applications
Training RNN
Unfolding of RNN
Architecture of RNN

Deep learning

Regularization for deep learning
Hyper parameters tuning
Abstractions of features using deep layers
Introduction to end to end learning
Drop out

Convolution networks with deep learning

CNN
RCNN
Faster RCNN
CNN with fully connected networks
Variants of pooling functions
Pooling

Autoencoders with deep learning

Applications of autoencoders (1)
Variational autoencoders
Applications of autoencoders
Undercomplete autoencoders (2)
Manifold learning with autoencoders

Generative deep learning models

GAN
Applications of GAN
Boltzmann machine
Deep belief machines
Restricted Boltzmann machine

Instructors

BITS Pilani Frequently Asked Questions (FAQ's)

1: Can I defer my programme if I am unable to devote a sufficient amount of time in the Post Graduate Programme in Artificial Intelligence and Machine Learning programme?

The candidates are allowed only once to defer from this batch to the next batch. The candidates will have to pay 10 % of the deferral fee. The candidates will be approved for the next immediate batch only after paying the deferral fee. For more information, the candidates can check the FAQ section of the website.

2: Is there any refunding scheme for this programme?

The programme offers refunding opportunities to its candidates if the candidates opt to leave this programme within 14 days of the course start date. The candidates will have to fill up the refund form given by the Admission cell. The candidates can hope to get a full programme fee.

3: What are the exam options provided in this programme?

This programme offers two exam options to which the candidates can choose according to their preferences- the first option is through online mode and the second option is through designated centres. The designated examination centres are – Bangalore, Chennai, Hyderabad, Pune, Mumbai, Goa, Delhi NCR, Pilani, Kolkata and also Dubai.

4: Does this programme offer job assistance?

This programme is designed to shape the candidates with industry-based practical knowledge. The Post Graduate Programme in Artificial Intelligence and Machine Learning certification course is equipped with exposing the candidates to experts from this field. This programme doesn’t provide placement assistance but the skills will help them gain a job directly after completion.

5: What is the required weekly effort in this programme?

The candidates will have to devote 8-10 hours of weekly effort to this programme. The programme is online-based and each candidate will have to devote weekly hours for the course to be completed on time.

6: Who will clear my doubts and queries in this course?

The participants can clear their doubts and queries about this programme in many ways. The participants can ask their questions directly to course instructors through live sessions. The participants can discuss the programme in discussion forums. In the discussion forum, BITS Pilani faculty members and teaching assistants will be there.

7: What will I receive along with the certification on successful completion?

On successful completion, the candidates will receive the certification offered by BITS Pilani. Along with it, the candidates will be handed out an Official transcript and programme GPA.

8: Is BITS Pilani UGC approved?

BITS Pilani is an institution approved under UGC regulations 2017. This programme is under Work Integrated Learning Programmes offered by BITS Pilani.

9: What is the Capstone Project?

The Capstone Project is an 8-week long project work that will expose the candidates to real-life problems in AI & ML. Throughout the project, they will be guided or mentored by faculty members and senior industry practitioners.

10: Will the candidates acquire any special status in the Post Graduate Programme in Artificial Intelligence and Machine Learning training?

The candidates will acquire BITS Pilani’s alumni status which is itself very special. It is special because it will mark the candidate as a member of the global elite community of BITS Pilani.