JECRC University

Faculty of Engineering

B. Tech. (CSE) with Specialization in Artificial Intelligence and Machine Learning

(In Association with Xebia)

Semester Subjects Offered L P

Semester 1 Foundation of Artificial Intelligence & Machine Learning 3
Semester 2 Python Programming 3
Semester 3 Supervised Learning in Machine Learning 3 1
Semester 4 Unsupervised Learning & Neural Networks 3 1

Semester 5 Advance Neural Network & Deep Learning 3 1

Semester 6 Natural Language Processing 3 1

Semester 7 Search Algorithms 3 1
Computer Vision & Image Processing 3 1
Semester 8 Reinforcement Learning 3 1

Total Credits: 34
Total Duration: 492

1
 Semester 1 Total Credits: 3
Lecture: 3 Credits Practical:

FOUNDATIONS OF ARTIFICIAL INTELLIGENCE & MACHINE LEARNING

UNIT I Foundations of AIML

Introduction and history of AI, Sources of Data, Introduction to AI, A Brief History of Data
Science, Introduction to Machine Learning, History of AI, categories of ML systems, An
Introduction to AI, ML and AI Overlap with Each Other? Applications of ML, Types of Data,
Organization of Structured Data, Examples of Structured Data, Expansion of Structured Data,
What is Semi-structured Data?
UNIT II Basic Statistical Concepts
Introduction to Statistics, Classification of Statistical Methods, Descriptive Statistics,
Inferential Statistics, Scale of Measurements (Nominal, Ordinal, Ratio and Interval), Nominal
Scales, Nominal Scales, Ratio Scales, Mean, Median, Mode, Measures of Variability/Spread,
range, Quartiles and Interquartile Range, Standard Deviation (SD), Measures of Shape,
Skewness, Kurtosis
UNIT III Probability Theory
Principles of Counting, Introduction and Definitions of Probability Theory, Conditional
Probability, Bayes Theorem, Discrete Probability Distribution, Covariance and Correlation,
Continuous Probability Distribution, Central Limit Theorem, Hypothesis Testing
UNIT IV Matrices
Introduction to Matrices, Matrix Notations and Types, Matrix Equality, Operations on
Matrices, Determinants, Singularity of a Matrix, Orthogonal Matrix, Elementary
Transformations and elementary matrices, Echelon forms and echelon transformations,
Matrix Rank and Normal Form of a matrix, Vector Spaces and the axioms, Linear Dependence
and Independence of vectors, Consistency of linear system of equations, Eigenvalues and
eigenvectors, Cayley Hamilton Theorem, Linear Transformation and Orthogonal
transformation, Matrix Factorization and Types
UNIT V Linear Algebra
Introduction to Linear Algebra, Notations in Linear Algebra, Important Concepts of Linear
Algebra, Definitions of Linear Algebra, Introduction to mathematical modeling, Applications
of mathematical modeling, Principles and stages involved in developing a mathematical
model, Classification of mathematical modeling, Conceptualizing a mathematical model,
Concept of boundary conditions

2
 Semester 2 Total Credits: 4
Lecture: 3 Credits Practical: 1 Credits

PYTHON PROGRAMMING

UNIT I Setting up the Python Environment

Compiler vs. Interpreter, Statically vs. Dynamically Typed Languages, Introduction to Python,
Installing Python, Anaconda, Jupyter Notebook, Spyder, Components and Versions of Python,
Difference between Python 2 and Python 3, Python Distributions
UNIT II Programming with Python
Python REPL,Variables, control structures, functions and objects, First-class functions,
immutable data, strict and non-strict evaluation, Recursion instead of an explicit loop state,
Functions, iterators, and generators, Writing pure functions, functions as first-class objects,
Using strings, tuples and named tuples, Using lists, dicts, and sets, The itertools module,.Best
practices and clean coding,.Reading data files into Python, writing files,.Introduction to
Python libraries
UNIT III Data Preprocessing
Introduction,Introduction to Pandas and Basic Concepts of Pandas, Data Cleaning and
Preparation, Handling Missing Data, Filtering out Missing Data, Filling in Missing Data, Data
Transformation, Removing Duplicates, Transforming Data Using a Function or Mapping,
Replacing Values, Renaming Axis Indexes, Discretization and Binning,Detecting and Filtering
Outliers, Permutation and Random Sampling, String Manipulation, Feature Engineering
Unit IV Statistical Modeling
Derived Variables, Basic Exploratory Data Analysis, Methods for EDA and Examples, Statistical
Modeling, Curve Fitting: Linear Regression, Nonlinear Regression

5
Python Programming Lab

List of Programs

1. Setting up the Python Environment using Anaconda IDE : Know Jupyter & Spyder
2. Write a program to perform Functions in Python
3. Write a program in Python First Class Functions & Immutable Data
4. Write a program in Python exploring Iterators
5. Write a program in Python exploring Generators
6. Work in Python using Collections
7. Write a program in Python to perform Higher Order Function-I
8. Write a program in Python to perform Higher Order Functions-II
9. Write a program in Python to perform File Operation in Python
10. Write a program in Python to perform Data Preprocessing
11. Write a program in Python to perform Exploratory Data Analysis
12. Write a program in Python to actuate Curve Fitting
Course Learning Outcomes (CLOs):-
On completion of this course, the students will be able to:-
1. Get a hands-on programming experience.
2. Learn statistical implementations
3. Carry out exploratory data analysis

6
 Semester 3 Total Credits: 4
Lecture: 3 Credits Practical: 1 Credits

SUPERVISED LEARNING

UNIT I: Difference Between Supervised and Unsupervised Learning

Machine learning,why we need machine learning, machine learning process State the different types
of learning: Supervised,unsupervised and reinforcement learning,, Detailing out on labeled data and
its types,classification and regression models, unlabeled data and its types,clustering model; Gradient
Descent- Overview,Gradient Descent, Finding a Minimum Using Gradient Descent, Estimating the
Gradient, Using the Gradient Descent, Example, Loss Function, Different Loss Functions,

UNIT II: Regression Techniques

Regression Technique, Origin of Regression, Regression in Real World, regression concepts, Regression
Types, Linear Regression Types, Linear Regression Variance, Co-Variance, Linear Regression
Correlation Coefficient, OLS, R Squared, Goodness of fit, Linear Regression Using Gradient Descent,
Gradient Descent Explained with an Example, Stochastic Gradient Descent, Cost Function –Partial
Derivative, Testing Model Using Cross Validation, Cross Validation Types, regularized regression, Ridge
Regression, lasso regression, L1 vs L2 Norm – Regression, Generalized Linear Regression, RANDOM
COMPONENT OF A GLM

UNIT III: Classification Techniques- Decision Tress

Classification Technique, Decision Tree, Decision Tree Illustration using Sample Dataset, concept of
homogeneity., entropy, Entropy Explained with Rainfall Example, plot of entropy versus the
proportionsm, Information Gain, Algorithms to Create a Decision Tree, Gini Index, Truncation and
Pruning, Decision Tree Working Methodology, Decision Tree Tuning Parameters

UNIT IV: Classification Techniques- Naïve Bayes

Naïve Bayes, bayes theorem., Example, Naïve Bayes Algorithm for Categorical Data, Popular Naive
Bayes Classifiers, Types of Naive Bayes Classifier, Naïve Bayes for Text Classification, popular naive
bayes classifiers, Naïve Bayes Algorithm, K Nearest Neighbour classification , Curse of Dimensionality,
K-Factor, Implementation of KNN using Python

UNIT V Ensemble Methods

Ensemble Methods ,Why Ensemble?, Example, Methods for Constructing Ensemble, advantages and
disadvantages of ensembling. Random Forest, Random Forest Example, Random Forest Use Case,
Random Forest Algorithm, Comparing other Models Accuracy, Bootstrapping and Bagging, Out of Bag
Error, OOB Score Before Tuning, OOB and Hyper Parameter Tuning, Ensemble Model Using Majority
Voting, Gradient Boosting, Weak Learner, Gradient Boosting Example, Moving towards XGBoost,
Parameters of XGBoost

7
Supervised Learning Lab

List of Programs
1. Understanding the Learning Implementation on Jupyter Notebook
2. Using NumPy functions in Jupyter
3. Using Pandas in Jupyter
4. Using SciPy in Jupyter
5. Using Simple Linear Regression, calculate Gradient and Cost minimum, Along with line
of best fit.
6. Understand Linear Regression and other regression techniques using house prices
prediction dataset.
7. Understanding Decision tree with sample dataset.
8. KNN algorithm explained with Cancer Data. (Using Python)
9. Identifying optimal K value in K-means Clustering algorithm. (Using Python)
10. Random Forest algorithm explained with classification and Regression (Using Python)
11. Implement Gradient Boosting Machine using Python.
12. Understand Logistic Regression model using Iris dataset (using Python)

Course Learning Outcomes (CLOs):

On completion of this course, the students will be able to:-
1. Understand the supervised Learning Concepts.
2. Learn the SL Techniques

8
 Semester4 Total Credits: 4
Lecture: 3 Credits Practical: 1 Credits

UNSUPERVISED LEARNING & NEURAL NETWORKS

UNIT I Dimensionality Reduction

Introduction, Singular Value Decomposition, SVD code:, Principal Component Analysis (PCA),
Isometric Maps (Isomaps), Multidimensional Scaling (MDS), ISOMAPS with MDS, ISOMAPS
(Code), Visualizing the ISOMAPS Data, Applying PCA on the Same Data, Visualization of PCA,
Feature Selection Techniques, Wrapper Method
UNIT II Clustering
What is Clustering and Why is it Important?, Techniques in Clustering, K-Means Clustering,
Steps for K-Means Algorithms, Density Based Spatial Clustering (DBSCAN), Types of Points in
DBSCAN, DBSCAN Example, DBSCAN: Advantages, DBSCAN: Disadvantages, Hierarchical
Clustering, Dendrograms, Hierarchical Clustering Code, DBSCAN Dendrogram Visualization
UNIT III Neural Networks
Introduction to Neural Networks, Types of Neural Networks, Perceptron, Limitations of
Perceptron, Activation Functions, Types of Activation Functions, Linear Activation Function,
Non- Linear Data, Non-Linear Activation Function (Sigmoid ), Non- Linear Activation Function
(TanH), Non-Linear Activation Function (ReLu), Non-Linear Activation Function (Leaky ReLu),
Derivative of Activation FunctionsNeural Networks, Feed Forward Network, ANN Forward
Propagation, Flow of Data in ANN, Backpropagation, Cost Function in Backpropagation, ANN
Evaluation, Complete Flow of Data in Neural Network, ANN Training ANN Design, Dropout in
Neural Networks
UNIT IV Understanding Images:
Understanding Images, Need of Convolution Neural Network Convolution, Neural Network
Working, Working of CNN with Kernel, Understanding Convolution Mathematically, An
Example of CNN, Convolution of Images, Convolution Neurons Visualization, Parameters for
Feature Maps, Activation Function in Convolution Neural Network, Pooling Step, Advantages
of Pooling, Batch Normalization, Typical Convolution Neural Network, Training CNN using
Backpropagation, Steps for CNN Backpropagation, Example of Convolution Neural Network
Architecture, Visualization of Convolutional Neural Networks,
UNIT V RNN Overview
Using MLP instead of RNN, Recurrent Neural Network (RNN), Steps in Recurrent Neuron, RNN
Mathematically, Example of Forward Propagation for RNN, Back Propagation in Recurrent
Neural Network, Steps for Back Propagation, Applications of RNN, Limitations of RNN, LSTM
Conveyor Belt Analogy, Architecture of LSTM, Gates in LSTM, Forget Gate, Input Gate, Output
Gate

Unsupervised Learning & Neural Networks Lab

9
List of Programs
1. Write a program for using PCA on MNIST Dataset.
2. Write a program for using PCA on Cat and Dog Dataset.
3. Write a program for using LDA on Cat and Dog Dataset.
4. Write a program for using DBSCAN on IRIS Dataset.
5. Write a program for using SVD on MNIST Digits Dataset.
6. Write a program for Feature Selection Techniques (Forward/ Backward).
7. Write a program for K-Means Clustering on IRIS Dataset.
8. Write a program for Hierarchical Clustering on Customers Dataset.
9. Write a program for Neural Networks on Mobile Price Classification.
10. Write a program for Convolution Neural Network on MNIST Dataset.
11. Write a program for Convolution Neural Network on Malaria Dataset.
12. Write a program for Convolution Neural Network on Aerial Cactus Dataset.

Course Learning Outcomes (CLOs)

On completion of this course, the students will be able to:-
1. Understand Unsupervised Techniques
2. Understand Neural Network basics with hands-on.

10
 Semester 5 Total Credits: 4
Lecture: 3 Credits Practical: 1 Credits

ADVANCED NEURAL NETWORKS & DEEP LEARNING

UNIT I: RNN
Gradient Descents, Recurrent Neural Network, Predicting the next character using RNNs,
Hopifi eld Network, Gated Recurrent Unit (GRU), Bidirectional RNN
UNIT II Deep Learning:
Introduction to Deep Learning, Introduction to Tensorflow, creating a Deep Learning Network
using Tensorflow
UNIT III Boltzmann Machines
Introduction to Boltzmann Machines, Restricted Boltzmann Machines, Collaborative Filtering
using Boltzmann Machines, Markov Random Fields, Deep Boltzmann Machine
UNIT IV Deep Belief Networks
Introduction to Deep Belief Network, Stacking RBM to create Deep Belief Network, Wake
Sleep Algorithm
UNIT V Modern Statistical Concepts
Learn about confidence intervals, define jackknife regression, Explain graphical models,
Describe tetter goodness of fi t and yield metrics

11
Advanced Neural Networks Lab

List of Programs

1. Build a perceptron model from Scratch

2. Write a program to visualize different activation functions and their derivative
3. Write a program for Hyperparameter Tuning and Optimization in Tensorflow
4. Write a program for simulation of Jackknife estimation of mean and median
5. Write a program for understanding different tensorflow syntax and different
operations.
6. Write a program to understand Keras in Tensorflow
7. Write a program for Linear Regression in Tensorflow
8. Write a program for Logistic Regression with Tensorflow
9. Write a program for Next character prediction using RNN in Tensorflow
10. Write a program for next character prediction using Bidirectional RNN in ensorflow
11. Write a program for next word prediction using RNN in Tensorflow
12. Write a program for Collaborative Filtering using RBM in Tensorflow
13. Write a program for Classification using DBN
14. Write a program for A/B Testing using Bayesian Method in Tensorflow

Course Learning Outcomes (CLOs)

On completion of this course, the students will be able to:-
1. UnderstandRegression with Tensorflow
2. Learn Keras concepts
Understand Collaborative Filtering

12
 Semester 6 Total Credits: 4
Lecture: 3 Credits Practical: 1 Credits

NATURAL LANGUAGE PROCESSING

UNIT I Natural Language Processing

Introduction to Natural Language Processing
Types of NLP systems, How computer understands text, Terminologies used in NLP, Steps
Involved in NLP, Steps involved in preprocessing , Pipeline of NLP Problems o Challenges in
NLP
UNIT II Words & Vectors
Concepts of words and vectors, Techniques of converting words to numbers, GloVe Word
Embeddings, Word2Vec and its types, such as Skip Gram, Model and Continuous BOW o
Advanced word vectors, limitations of CBOW and Skip Gram
UNIT III Processing Techniques
Word window classification, Dependency parsing, Constituency parsing o Machine
translation, Attention, End to end models for speech processing, Deep learning for speech
recognition, Tree recursive neural networks o RNN for language modelling, Dynamic neural
network for question answering
UNIT IV: Case Studies
Smart Home Services Provider Uses Natural Language Generation to Create Highly
Personalized Website Copy, Online Education Company Improves Customer Support with
Autosuggestion of Macros, Using Natural Languag for Health care Summaries, Microsoft Gets
the Pulse of Customer Sentiment with Natural Language Processing

13
Natural Language Processing Lab
List of Programs
1. Text Classification using Word Embeddings.
2. Find Synonyms and antonyms using Word Embeddings.
3. Introduction to Topic Modelling.
4. Converting a Foreign Language to English using Machine Translation(German to
English).
5. Twitter Sentiment Analysis.
6. Explaining Lemmatization, PoS Tagging,
7. Stemming and Tokenization using an Example.
8. Understanding Dependency Parsing in a given sentence.
9. Perform Speech to Text Conversion using PyAudio and Google Speech Recognition.
10. Creating Custom Speech Recognition Corpus.
11. Introduction to Dynamic Memory Network.
12. Dialog Generation using Deep Learning.

Course Learning Outcomes (CLOs):

On completion of this course, the students will be able to:-

1. Understand the sentiment analysis
2. Learn text classification concepts

14
 Semester 7 Total Credits: 4
Lecture: 3 Credits Practical: 2 Credits

SEARCH ALGORITHMS

Unit I: Problem Solving and Search Strategies

State space representation of AI problems, Considerations for Problem Solving, Introduction
to uninformed and informed search, Random Search, Search with Closed and Open List,
Problem Solving as Search, Data Driven or Goal Driven Search, Generate & Test., Properties
of Search Methods
Unit II: Search Algorithms

Depth First Search, Disadvantages of Depth First Search,Breadth First Search, Disadvantages
of Breadth first search, Traversing a Maze, Searching for a Gift, Implementing Depth First
Search & Breadth First Search, Web Spidering, Depth First Iterative Deepening, Advantages
of Iterative deepening, Uniform-cost search: The Dijkstra's algorithm, Disadvantages of
Uniform cost search
Unit III: Heuristic Search

Introduction to Heuristic Search, Choosing a good heuristic, The 8-Puzzle, Monotonicity

Modified Travelling Salesman Problem,Best Fit Search, A* Algorithm, Iterative deepening A*,
Generalization of Problems, Optimization
Unit IV: Hill Climbing

Hill Climbing Algorithm, Steepest Ascent Hill Climbing, Stochastic Hill Climbing, Disadvantages
of Hill Climbing - Foothills, Plateaus & Ridges

15
Search Algorithms Lab
List of Programs

1. Perform DFS using Python

2. Perform BFS using Python
3. Determine whether goal- or data-driven search would be preferable for solving each
of the following problems. Justify your answer.
a. Diagnosing mechanical problems in an automobile.
b. You have met a person who claims to be your distant cousin, with a common
ancester named “John Doe”. Verify the claim.
c. Another person claims to be your distant cousin. He does not know the
ancester’s name, but knows that it was more than eight generations back. You
would like to find this ancester or determine she does not exist.
4. A snake is in a maze of N*N sized matrix. It has to travel from source to destination
block, but can move only forward or down. Ow will you achieve this. Use Python.
Note: Grey blocks are deadends.

5. Given a graph and a source vertex in the graph, find shortest paths from source to all
vertices in any given graph. Use Dijkstra’s algorithm.
6. What is mother vertex? Find a mother vertex in any given graph.
7. Solve the traveling salesman problem using genetic algorithm in python.
8. Imagine a real life situation, where you are using a Map (Map on a paper). You are at
point A and want to reach point B. Which algorithm you would choose and why.
9. Implement question 8, using python.
10. Explain a use case and implement Stochastic hill climbing algorithm using python.
11. Solve the 8 puzzle problem using appropriate algorithm and python.
12. Assume you have 10 given search result. Find the best solution for the keywords you
select while searching on your python console.

Course Learning Outcomes (CLOs):

On completion of this course, the students will be able to:-

1. Understand the image Transformation Techniques
2. Learn image Segmentation using Python

16
 Semester 7 Total Credits: 4
Lecture: 3 Credits Practical: 1 Credits

COMPUTER VISION & IMAGE PROCESSING

Unit I: Introduction to Computer Vision and Image Processing

Image Processing, Elements of Image Processing System, Computer Vision, Computer Graphics,
Application Areas, Imaging Geometry, Image Sampling, Mathematical Tools, Image transformations:
2D and 3 D Transformation, Image Enhancements-Intro, Image Segmentation-Intro, Cognitive Aspects
of Color, VR/AR, Object Recognition, Object Tracking

Unit II: Introduction to Open CV

Introduction, GUI Features, Operations: Pixel Editing, Geometric Transformations, Feature Detection,
Videon Analysis and Tracking, Stereo Imaging, Caliberation, OpenCV-Python, Image Denoising, Object
Detection, Transformation and Spatial Filtering Introduction, Functions, Histogram, Histogram
Equalization, Histogram Matching (Specification), Local Histogram Processing, Using Histogram
Statistics for Image Enhancement Introduction to Spatial Filtering, Smoothing & Sharpening Image
Fileters

Unit III: Image compression

Fundamentals, Coding Redundancy, Spatial Redundancy, Irrelevant Information, Models,

Compression Methods, Huffman Coding, Golomb Coding, Arithmetic Coding, LZW Coding, Run-Length
Qlding, Symbol-Based Coding, Bit-Plane Coding, Block Transform Coding, Predictive Coding, Wavelet
Coding

Unit IV: Image Segmentation

Human Vision, Applications, Point Line and Edge Detection, Shot Boundary Detection, Interactive
Segmentation, Image Segmentation by Clustering Pixels, Basic Clustering Methods, The Watershed
Algorithm, Segmentation Using K-means, Graphs, Fitting, Motion Segmentation, Model Selection,
Case Studies

Unit V: Object Recognition and tracking

Introduction, Shape correspondence and shape matching, Sliding Window Method, Patterns,
Structural Methods, Deformable Objects, Tracking, Strategies, Matching, Tracking with Filters, Data
Association, Particle Filtering

17
Computer Vision & Image Processing Lab
List of Objectives

1. Discuss about important transformations used in imaging.

2. Write a code to detect car from image.
3. Write a program that will detect a change in the video feed.
4. Explain the 4 variations of image denoising in OpenCV, using Python.
5. How to perform histogram matching using OpenCV?
6. Compress an image without losing the quality of the image.
7. Analyze an image using histogram and enhance the image using Histogram statistics.
8. Detect Lane lines from images or video, primarily by using edge detection.
9. Implement image segmentation using edge detection segmentation and python.
10. Implement image segmentation using clustering and python.
11. Implement image segmentation using CNN.
12. Write a program to calculate the distance from an object to the camera using stereo
vision.
13. Perform motion estimation in videos.
Course Learning Outcomes (CLOs):

On completion of this course, the students will be able to:-

 Understand the image Transformation Techniques
 Learn image Segmentation using Python

18
 Semester 8 Total Credits: 4
Lecture: 3 Credits Practical: 1 Credits

REINFORCEMENT LEARNING
Unit 1: Introduction to Reinforcement Learning

Introduction to Reinforcement Learning, Reinforcement Learning Overview, Need of RL, Elements of

RL, Exemplary explanation, Origin and Overview, Challenge of Reinforcement Learning, Evaluative
Feedback, Evaluation & Instruction, Incremental Implementation, Relation with other fields

Unit 2: Action Value Methods

SARSA, Explore & Exploit Dilemma, Multi-armed Bandits Action Value Methods, k-armed Bandit
Problem, The 10-armed Test Bed, Tracking a Nonstationary Problem, Optimistic Initial Values,
Reinforcement Comparison, Pursuit Methods, Associative Search, Gradient Bandit Algorithms, Upper-
Confidence-Bound Action Selection. The Reinforcement Learning Problem, Thompson’s Sampling

Unit 3: Agent- Environment Interface

Agent- Environment Interface, Goals & Rewards, The Markov Property, Markov Decision Processes,
Value Functions, Returns & Episodes, Finite Markov-Decision, Optimality & Approximation Markov
Decision Process, Markov chains, Markov reward process (MRP). Introduction to and proof of Bellman,
Bellman equations in MRP.

Unit 4: Dynamic programming

Dynamic programming Policy Evaluation, Improvement, Iteration, Value Iteration, Asynchronous DP,
Generalized Policy, Efficiency of Dynamic programming, Prediction and Control by Dynamic
Programing

Unit 5: Monte Carlo Methods

Monte Carlo Methods MonteCarlo Prediction, Estimation of Action Values, Monte Carlo Control, Off-
Policy Prediction, Inremental, On policy and off policy learning, Importance sampling, TD Prediction,
Optimality of TD, Actor Critic Method, Temporal Difference Learning, Q-Learning, Approximation
Methods for prediction

19
Reinforcement Learning Lab
List of Objectives

1. Write a program to demonstrate Action Value Methods

2. Write a program to actuate 10-armed Test Bed.

3. Perform Associative Search in RL

4. Computer the return agent for a particular action and choosing the best

5. Build a Stock Reinforcement Learning Algorithm

6. Apply RL in Large Discrete Action Spaces

7. WAP to analyse Bellman Equation for Value Function

8. Write a program to Apply the MCMC Method in RL

9. Apply Monte Carlo Method to solve a problem by generating suitable random numbers

10. Solve a prediction Problem using TD Prediction

20
21