0% found this document useful (0 votes)
779 views73 pages

M.Tech ArtificialIntelligence and Data Science-SYLLABUS

The document outlines the scheme of instruction and syllabus for the M. Tech Programme in Artificial Intelligence and Data Science offered by APJ Abdul Kalam Technological University. The program is structured over 4 semesters and requires students to complete a total of 68 credits. The courses include both core and elective subjects focused on artificial intelligence and data science topics, along with labs, seminars, and a final year project.

Uploaded by

Maya M S
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
779 views73 pages

M.Tech ArtificialIntelligence and Data Science-SYLLABUS

The document outlines the scheme of instruction and syllabus for the M. Tech Programme in Artificial Intelligence and Data Science offered by APJ Abdul Kalam Technological University. The program is structured over 4 semesters and requires students to complete a total of 68 credits. The courses include both core and elective subjects focused on artificial intelligence and data science topics, along with labs, seminars, and a final year project.

Uploaded by

Maya M S
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 73

APJ Abdul Kalam Technological University

Cluster 6: Ernakulam I

M. Tech Programme in
Arti cial Intelligence and Data
Science
Scheme of Instruction and Syllabus: 2020 Admissions

APJ Abdul Kalam Technological Universit


Page 1 of 73 Cluster 6 | M. Tech Programme in AI & DS

fi
y

APJ Abdul Kalam Technological University


(Cluster 6: Ernakulam I

M. Tech Programme in Arti cial Intelligence and Data Science


Scheme of Instruction

Credit requirement: 68 credits (23+19+14+12)


Normal Duration - Regular: 4 semester Maximum Duration - Regular: 6 semester
Courses - Core courses: Either 4 or 3 credits courses; Elective courses: All of 3 credit
Allotment of credits and examination schem

Semester I (Credits: 23

End Semester Exam


Exam Course Internal
Subjects L-T-P Credits
Slot Code Marks Marks Duration

A 06 DS 6 01 1 Linear Algebra and Optimization 4-0-0 40 60 3 hrs 4

B 06 DS 6 02 1 Stochastic Models and


4-0-0 40 60 3 hrs 4
Numerical Optimization

C 06 DS 6 03 1 Principles of Arti cial Intelligence


4-0-0 40 60 3 hrs 4
and Machine Learning

D 06 DS 6 04 1 Exploration and Statistical


3-0-0 40 60 3 hrs 3
Analysis for Data Science

E 06 DS 6 x5 1 Elective I 3-0-0 40 60 3 hrs 3

F 06 DS 6 06 1 Research Methodology 0-2-0 100 2

S 06 DS 6 07 1 Seminar 1 0-0-2 100 2

U 06 DS 6 08 1 Machine Intelligence Lab 0-0-3 100 1

23
Credits

Elective I

Course Code Subjects

06 DS 6 15 1 Ensemble Models

06 DS 6 25 1 Convolutional Neural Network

06 DS 6 35 1 Soft Computing

06 DS 6 45 1 Computer Vision

APJ Abdul Kalam Technological Universit


Page 2 of 73 Cluster 6 | M. Tech Programme in AI & DS

fi
)

fi
e

Semester II (Credits: 19

End Semester Exam


Exam Internal
Course Code Subjects L-T-P Credits
Slot Marks Marks Duration

A 06 DS 6 01 2 Big Data Analytics * 4-0-0 40 60 3 hrs 4

B 06 DS 6 02 2 Deep Learning and Arti cial


3-0-0 40 60 3 hrs 3
Neural Network

C 06 DS 6 03 2 Genetic Algorithms 3-0-0 40 60 3 hrs 3

D 06 DS 6 x4 2 Elective II 3-0-0 40 60 3 hrs 3

E 06 DS 6 x5 2 Elective III 3-0-0 40 60 3 hrs 3

F 06 DS 6 06 2 Mini Project 0-0-4 100 2

U 06 DS 6 07 2 Deep Learning Lab 0-0-3 100 1

19
Credits

Elective II

Course Code Subjects

06 DS 6 14 2 R for Data Science

06 DS 6 24 2 Data Analytics and Scalable Algorithms

06 DS 6 34 2 Scalable Systems for Data Science

06 DS 6 44 2 Knowledge Engineering and Data Science

Elective III

Course Code Subjects

06 DS 6 15 2 Big Data for Internet of Things

06 DS 6 25 2 Arti cial Intelligence and Robotics

06 DS 6 35 2 Natural Language Processing

06 DS 6 45 2 Machine Learning Models and Storage


Management

* - Subject common to M.Tech Data Science/M.Tech Arti cial Intelligence and Data Scienc

APJ Abdul Kalam Technological Universit


Page 3 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
fi
)

fi
e

Semester III (Credits: 14

End Semester Exam


Exam Internal
Course Code Name L-T-P Credits
Slot Marks Marks Duration

A 06 DS 7 x1 1 Elective IV 3-0-0 40 60 3 hrs 3

B 06 DS 7 x2 1 Elective V
3-0-0 40 60 3 hrs 3

S 06 DS 7 03 1 Seminar II 0-0-2 100 2

F1 06 DS 7 04 1 Project Phase I 0-0-8 50 6

14
Credits

Elective IV

Course Code Subjects

06 DS 7 11 1 Arti cial Intelligence in Cyber Security

06 DS 7 21 1 Game Theory in Arti cial Intelligence

06 DS 7 31 1 Image and Video Analytics

06 DS 7 41 1 Cloud Data Management

Elective V

Course Code Subjects

06 DS 7 12 1 Data Visualisation Techniques

06 DS 7 22 1 Social Network Analysis

06 DS 7 32 1 Text Mining

06 DS 7 42 1 Data Warehouse and Data Lakes

Semester IV (Credits: 12

Exam Internal End Semester Exam


Course Code Name L-T-P Credits
Slot Marks
Marks Duration
F2 06 DS 7 01 2 Project Phase II 0-0-21 70 30 12

12
Credits

APJ Abdul Kalam Technological Universit


Page 4 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
fi
)

APJ Abdul Kalam Technological University


Master of Technology – Course Plan

SEMESTER I

M. Tech Programme in
Arti cial Intelligence and Data Science

APJ Abdul Kalam Technological Universit


Page 5 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
y



Course Year of
Course Name L-T-P : C
Code Introduction

06DS6011 Linear Algebra and Optimization 4-0-0 : 4 2020

Course Objectives:

1. To visualise vectors in n-space which is useful in representing data

2. To learn handling of linear system of equations using matrix as a tool

3. To introduce Eigen values and Eigen vectors which are signi cant in dynamic problems

4. To introduce matrix decompositions methods that reduce a matrix into constituent parts
which make it easier to calculate more complex matrix operations

5. To study optimization algorithms with single and multi-variables for large datasets
SYLLABUS:
Vector spaces, subspaces, bases and dimensions, systems of linear equations, Linear
transformations, Isomorphism, Inner product, Orthogonality, Eigenvalues and Eigenvectors,
Matrix factorisations, Function optimization, Newton’s method.
Course Outcomes:


Students should be able to apply

1. Theory of vector space in representing data

2. Matrix operations in solving system of linear equations

3. Matrix decomposition in solving system of equations

Module Course Content Hours


I Vector Spaces: Vector Spaces, Subspaces- De nition and Examples, 

Linear independence of vectors, Bases and dimension, Linear Span, 6

Field-De nition


Vector space in Rn: System of linear equations, row space, Column 

space and null space. Four fundamental spaces, relation between rank 7
and nullity, consistency theorem, basis from a spanning set and
independent set.
INTERNAL TEST 1 (Module I)

II Linear transformations: General linear transformation, Matrix of 



transformation, Kernel and range, properties, Isomorphism, change of 7

basis, invariant subspace, Linear functional.


Inner Product: Real and complex inner product spaces, properties of 

inner product, length and distance, Cauchy-Schwarz inequality, 6
Orthogonality, Orthogonal complement, Orthonormal bases, Gram
Schmidt orthogonalisation
INTERNAL TEST 2 (Module II)

III EigenSpace: Properties of Eigen values and Eigen vectors , Eigen 



values, Eigen vectors, minimal polynomial, Diagonalization, Orthogonal 7

diagonalization, Jordan canonical form


Matrix Factorization: LU decomposition, QR Decomposition and 6
singular value decomposition

APJ Abdul Kalam Technological Universit


Page 6 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
y

fi
fi
IV Optimization: Conditions for local minimization-One dimensional 

Search methods:Golden search method, Fibonacci method, Newton’s 

Method, Secant Method, Remarks on Line Search Gradient-based 7

methods-introduction, the method of steepest descent, analysis of 

Gradient Methods, Convergence, Convergence Rate.


Analysis of Newton’s Method, Levenberg-Marquardt Modi cation, 

Newton’s Method for Nonlinear Least-Squares. Conjugate direction 6
method, Conjugate Direction Algorithm, Conjugate Gradient Algorithm
for Non-Quadratic Quasi Newton method.
END SEMESTER EXAM (All Modules)
References:

1. Gilbert Strang Linear Algebra and It’s Applications, 4th edition, Cengage Learning, 2006.

2. Stephen Boyd, Lieven Vandenberghe, Introduction to Applied Linear Algebra: Vectors,


Matrices, and Least Squares, Cambridge University Press, 2018

3. W. Keith Nicholson, Linear Algebra with applications, 4th edition, McGraw-Hill, 2002

4. I.N Herstein, Topics in Linear Algebra, Wiley Eastern, 1975.

5. S.Kumaresan, Linear Algebra : A Geometric Approach, Prentice-Hall of India, 2000.

6. Seymour Lipschutz, Marc Lipson, Schaum’s outline of linear algebra, 3rd Ed., Mc Graw Hill
Edn., 2017

7. Edwin K.P. Chong, Stanislaw H. Zak, An introduction to Optimization, Second edition,


Wiley,2013

8. Mohan C. Joshi and Kannan M. Moudgalya, Optimization: Theory and Practice, Narosa
Publishing House, New Delhi,2004

APJ Abdul Kalam Technological Universit


Page 7 of 73 Cluster 6 | M. Tech Programme in AI & DS
y

fi
Course Year of
Course Name L-T-P : C
Code Introduction

06DS6021 Stochastic Models and Numerical Optimization 4-0-0 : 4 2020

Course Objectives:


1. To study the basic concepts of the theory of stochastic processes

2. To familiarise the most important types of stochastic processes

3. For studying the various properties and characteristics of processes

4. To study the methods for describing and analysing complex stochastic models
SYLLABUS:
Random variables and events, distributions, inequalities and limits, Stochastic processes,
Exponential distribution, Markov chains, Discrete Time Markov chains, Continuous Time Markov
chains, Mathematical models of optimisation.
Course Outcomes:


Students should be able to

1. Carry out derivations involving conditional probability distributions and conditional


expectations

2. Apply Markov chains in real world scenarios.

3. Solve di erential equations for distributions and expectations in time continuous processes.

4. Apply optimization processes to generate best results for real world applications.
Module Course Content Hours
I Events, Measurability, Independence - Sample Spaces, Events,
Measures, Probability, Independence, Conditional probability, Bayes’
theorem
4
Random Variables - Functions of random variables, Sequence of
random variables, Bernoulli, Binomial, Geometric, Poisson; Uniform,
Exponential, Normal, Lognormal, Expectations, Moments and Moment 5
generating functions, Random Vectors - Joint and Marginal
distributions, Dependence, Covariance, Transformations of random
vectors.

Conditioning RVs - Conditional Distribution of a RV, Computing 4


probabilities and expectations by conditioning, IT Application: Time-to-
a-pattern for password security
INTERNAL TEST 1 (Module I)

APJ Abdul Kalam Technological Universit


Page 8 of 73 Cluster 6 | M. Tech Programme in AI & DS
ff
y

II Inequalities and Limits of Events, RVs, Distributions - Inequalities: 4


Markov, Chebyshev, Jensen, Holder, Convergence of RVs: Weak and
Strong laws, Central limit theorem, Distributions of extreme.

Introduction to Stochastic Processes (SPs): De nition and examples


of SPs, classi cation of random processes according to state space 4
and parameter space, types of SPs, elementary problems. Stationary
Processes: Weakly stationary and strongly stationary processes,
moving average and auto regressive processes.

Exponential Distribution and Poisson Process - Construction of


Poisson Process from Exponential Distribution, Conditional Arrival
Times. Normal Distribution and Brownian Process - Construction of 6
Brownian Process from Normal Distribution, Hitting Times and
Maximum Values, Finance Applications: Option Pricing and Arbitrage
Theorem
INTERNAL TEST 2 (Module II)

III Markov Chains - Markovian property and Transition probabilities, 3


Irreducibility and Steady-State probabilities.

Discrete-time Markov Chains (DTMCs): De nition and examples of


MCs, transition probability matrix, Chapman-Kolmogorov equations; 6
calculation of n-step transition probabilities, limiting probabilities,
classi cation of states, ergodicity, stationary distribution, transient MC;
random walk and gambler’s ruin problem.

Continuous-time Markov Chains (CTMCs): Kolmogorov-Feller


6
di erential equations, in nitesimal generator, Poisson process, birth-
death process, stochastic Petri net, applications to queueing theory
and communication networks.
IV Mathematical modeling of optimization - Objective function,
Continuous functions and discrete functions, unimodal, convex, and
concave functions, Optimization constraints - internal and external.
Hessian matrix, Gradient-free search, Saddle point, Linear 10
programming and simplex method. Case Study : Transportation
problem  and Assignment problem, Case Study : Genetic Algorithms as
optimization problems.
END SEMESTER EXAM (All Modules)
References:
1. S.M. Ross, “Introduction to Probability Models”, 11th edition, Academic Press, 2014.

2. S.M. Ross, “Stochastic Processes”, 2nd Edition, John Wiley & Sons, 1996.

3. S. Resnick, “Adventures in Stochastic Processes”, Birkhauser, 1994.

4. A. M ller and D. Stoyan, “Comparison Methods for Stochastic Models and Risks”, John
Wiley & Sons 2002.

5. R.E. Barlow and F. Proschan, “Mathematical Theory of Reliability”, 1965.

6. J. Medhi, Stochastic Processes,3rd Edition,NewAge International, 2009.

7. S Karlin and H M Taylor, A First Course in Stochastic Processes, 2nd edition, Academic
Press, 1975.

8. Rao S. S., Optimization Theory and Applications, Wiley Eastern, 1984.

APJ Abdul Kalam Technological Universit


Page 9 of 73 Cluster 6 | M. Tech Programme in AI & DS
ff

fi
fi
fi
y

fi
fi
Course Year of
Course Name L-T-P : C
Code Introduction

06DS6031 Principles of Arti cial Intelligence and Machine Learning 4-0-0 : 4 2020

Course Objectives:

1. To familiarise basic principles of Arti cial Intelligence

2. To learn and design intelligent agents

3. To familiarise the basic areas of arti cial intelligence including problem solving, knowledge
representation, reasoning, decision making, perception and action

4. To master the fundamentals of machine learning, mathematical framework and learning


algorithm
SYLLABUS:
Intelligent Agents, Problem solving and search, Uninformed search, Knowledge and reasoning,
Probabilistic reasoning, Bayesian networks and decision theory, Neural networks, Issues in ANN
training, Types of ANN architectures, SVM.
Course Outcomes:


Students should be able to

1. Understand formal methods of knowledge representation, logic and reasoning

2. Understand foundational principles, mathematical tools and program paradigms of arti cial
intelligence

3. Understand the fundamental issues and challenges of machine learning: data, model
selection, model complexity

4. Analyse the underlying mathematical relationships within and across Machine Learning
algorithms and the paradigms of supervised and unsupervised learning

5. Apply intelligent agents for Arti cial Intelligence programming techniques


Module Course Content Hours
I Introduction- Intelligent Agent, Structure of Intelligent Agent
and  Environment. Problem solving and search strategies - Problem 7
solving agents, Problem-solving through Search - forward and
backward, state-space, blind, heuristic, problem-reduction, minimax,
constraint propagation, neural and stochastic;

Uninformed search strategies-Depth First Search, Breadth First


Search, Depth limited search, iterative deepening depth rst search,
bidirectional search. Informed search strategies-Best First Search, A*, 6
AO*.
INTERNAL TEST 1 (Module I)

II Knowledge and reasoning: A knowledge based agent, representation, 8


Propositional Logic-First Order Logic-Soundness and Completeness
-Forward and Backward chaining-Resolution-semantic networks-
Handling uncertain knowledge– Probabilistic Reasoning –making
simple and complex decisions.

Bayesian networks; Basics of decision theory, sequential decision


problems. 4

INTERNAL TEST 2 (Module II)

APJ Abdul Kalam Technological Universit


Page 10 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
y

fi
fi
fi
fi
fi
III Neural Networks: Introduction, Basic Architecture of Neural Networks,
7
Single Computational Layer:The Perceptron, Choice of Activation
functions, Number of Output Nodes and Loss Functions; Multilayer
Neural Networks, Training a Neural Network with Backpropagation.

Practical Issues in Neural Network Training: Problem of Over tting,


Vanishing and Exploding, Gradient Problems, Di culties in
Convergence, Local and Spurious Optima, Computational Challenges. 6

IV Types of Neural Architectures: Simulating Basic Machine Learning


with Shallow Models, Radial Basis Function Networks, Restricted 7
Boltzmann Machines, Recurrent Neural Network: Architecture, Training,
Applications. Convolutional Neural Network: Architecture, Training,
Applications-Confusion Matrix, Precision, Recall, F Measure.

Support Vector Machine: Architecture, Training, Applications.


Parameter Estimation Bias -Mean Squared Error -Relative E ciency – 5
Standard Error - Maximum Likelihood Estimation.
END SEMESTER EXAM (All Modules)
References:


1. Stuart Russel, Peter Norvig, Arti cial Intelligence ‘A modern Approach, Prentice Hall PTR.

2. Elaine Rich and Kevin Knight. Arti cial Intelligence, 3e, Tata McGraw Hill, 2017

3. Charu C. Aggarwal, Neural Networks and Deep Learning, Springer

4. Earl Gose, Richard O Duda, Peter E.Hart, David G.Stork, Pattern Recognition, PHI
Learning

5. Richard O Duda, Peter E.Hart, David G.Stock, Pattern Classi cation, Wiley ,India,
Second Edition.

APJ Abdul Kalam Technological Universit


Page 11 of 73 Cluster 6 | M. Tech Programme in AI & DS
y

fi
fi
ffi
ffi
fi
fi
Course Year of
Course Name L-T-P : C
Code Introduction

06DS6041 Exploration and Statistical Analysis for Data Science 3-0-0 : 3 2020

Course Objectives:

1. To obtain a Comprehensive knowledge of various tools and techniques for Data


transformation and visualisation

2. To learn the probability and probabilistic models of data science

3. To learn the basic statistics and testing hypothesis for speci c problems

SYLLABUS:
Data Science process, Memorization methods, Unsupervised models, Univariate data
exploration, Data visualisation, Prediction and ltering, Probability theory and Statistics.

Course Outcomes:


Students should be able to

1. Apply exploratory data analysis and create insightful visualisations to identify 



patterns

2. Understand the statistical foundations of data science and analyse the degree 

of certainty of predictions using statistical test and models

3. Understand the basic probability principles and techniques


Module Course Content Hours
I Data Science process - Roles and stages in a data science project,
Working with les and databases, Exploring and managing data.
5

Exploratory Data Analysis, Selecting models for data science problems,


Evaluation and validation of models, Documentation and deployment, 5
Presenting results to stakeholders.
INTERNAL TEST 1 (Module I)

II Memorization methods - single variable and multivariable models,


Linear and logistic regression. Unsupervised methods - Cluster 5
analysis, K-means algorithm, clusterboot method, Association rules.

Exploring Univariate Data - Histograms - Stem-and Leaf Quantile


Based Plots - Continuous Distributions -Quantile Plots- QQ Plot- Box 5
Plots.
INTERNAL TEST 2 (Module II)

III Python based data visualization, Prediction using linear regression -


single variable and multi-variable models, Collaborative ltering - user 9
based ltering and item based ltering.

APJ Abdul Kalam Technological Universit


Page 12 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
fi
y

fi
fi
fi
fi
IV Probability Concepts -Axioms of Probability - Conditional Probability
and Independence - Bayes Theorem - Expectation - Mean and 6
Variance Skewness Kurtosis; Common Distributions - Binomial Poisson 

Uniform - Normal Exponential Gamma-Chi-Square Weibull Beta


Introduction to Statistics - Sampling, Sample Means and Sample 

variance sample moments, covariance, correlation, Sampling
Distributions - Parameter Estimation Bias - Mean Squared Error -
Relative E ciency - Standard Error - Maximum Likelihood Estimation. 7
Comparing Two Samples - A/B Testing - ANOVA.
END SEMESTER EXAM (All Modules)
References:


1. Boris Lublinsky, Kevin T. Smith. Alexcy Yakubovich, "Professional Hadoop Solutions", Wiley,
2015
2. Jure Leskovec, Anand Rajaraman, Jeffrey D. Ullman, "Mining of Massive Datasets".
Cambridge University Press, 201
3. Nathan Yau, "Visualize This: The Flowing Data Guide to Design, Visualization and Statistics",
Wiley, 201
4. Nina Zumel, John Mount "Practical Data Science with R··. Manning Publications. 201
5. Sameer Madhavan , “Mastering Python for Data Science”, Packt Publishing Limited, 201
6. Tony Ojeda, Sean Patrick Murphy, Benjarnin Bengfort. Abhijit Dasgupta. "Practical Data
Science Cookbook", Packt Publishing Limited, 201
7. W. N. Venables. D. M. Smith and the R Core Team, "An Introduction to R", 201
8. Wendy L. Martinez, Angel R. Martinez, Computational Statistics Handbook with MATLAB,
Second edition, Chapman Hall/CRC, 2008
9. Douglas C. Montgomery, George C. Runger, Applied Statistics and Probability for Engineers,
Sixth Edition, Wiley, 2013
10. Dr.J.Ravichandran, Probability And Statistics For Engineers, First Edition,Wiley, 2010

APJ Abdul Kalam Technological Universit


Page 13 of 73 Cluster 6 | M. Tech Programme in AI & DS

ffi

Course Year of
Course Name L-T-P : C
Code Introduction

06DS6151 Ensemble Models 3-0-0 : 3 2020

Course Objectives:


1. To familiarise the various ensemble algorithms
2. To implement the algorithms and compare performance of each one
SYLLABUS:
Bias-variance tradeo , Non-generative models, Clustering, Combination methods, Diversity and
Pruning, Learning methods, Applications.
Course Outcomes:


Students should be able to
1. Understand the various applications of AdaBoost, Random Forest and other related areas
2. Understand the basic concepts of rule extraction and combination methods
Module Course Content Hours

I Bias, Variance, and the tradeo , Ensemble learning, Di culties in


ensemble learning, Switching Net models.
4

Non-generative models - Voting, Stacking. Generative models -


Boosting, Bagging, Random Forests. AdaBoost algorithm, Multiclass
extension. Two ensemble paradigm, Random Tree ensembles, 7
Combination using averaging.
INTERNAL TEST 1 (Module I)

II Clustering - Consensus clustering, Use of OpenEnsembles. Similarity


based methods and graph based methods. Relabeling based and 6
transformation based methods.

Combination methods - Algebraic methods, Behaviour Knowledge


space method, Decision template method, Dynamic classi er selection, 4
Mixture of experts.

INTERNAL TEST 2 (Module II)

III Ensemble diversity, Error decomposition, Diversity measures - 6


pairwise and non-pairwise, Information theoretic diversity.

Ensemble pruning, Categorization of pruning methods, Ordering


based and clustering based pruning, Optimization based pruning. 5

IV Semi-supervised learning, Active learning, Cost-sensitive learning,


Class-imbalance learning, Reduction of ensemble to single model, Rule 7
extraction and visualization of ensembles.

Application : Classifying fraudulent transactions, Predicting Bitcoin 3


prices, Sentiment evaluation in Twitter
END SEMESTER EXAM (All Modules)

APJ Abdul Kalam Technological Universit


Page 14 of 73 Cluster 6 | M. Tech Programme in AI & DS
ff

ff

fi
ffi

References:


1. George Kyriakides, Konstantinos Margaritis, “Hands-on Ensemble Learning with Python”,


Packet Publishers

2. Alan Gelfand, Crayton Walker, “Ensemble Modeling”, Marcel Decker Inc, 1984.

3. Zhi-Hua Zhou, “Ensemble Models - Foundations and Algorithms”, CRC Press

APJ Abdul Kalam Technological Universit


Page 15 of 73 Cluster 6 | M. Tech Programme in AI & DS
y

Course Year of
Course Name L-T-P : C
Code Introduction

06DS6251 Convolutional Neural Network 3-0-0 : 3 2020

Course Objectives:


1. To familiarise how convolution networks are designed

2. To perform feature extractions and classi cations
SYLLABUS:
Pattern classi cation, Feature extraction, Convolutions, Types of layers, Visualisation of CNNs,
Applications, Ensembles.
Course Outcomes:


Students should be able to

1. Understand how to build a convolutional neural network, including recent variations such as
residual networks

2. Know how to apply convolutional networks to visual detection and recognition tasks
Module Course Content Hours
I Pattern classi cation - Linear classi er, Multiclass classi er, Linear
separability, Feature extraction, Arti cial Neural Networks - Activation,
10
Bias, Initialization, Convolutions and Pooling activities.

Introduction : LexNet, AlexNet


INTERNAL TEST 1 (Module I)

II Deriving convolution from fully connected layer, Role of convolution,


Backpropagation of convolution layers and pooling layers, Designing 8
and training ConvNets.
INTERNAL TEST 2 (Module II)

III Analyzing quantitative results from ConvNets, Other types of layers -


Local response normalization, Spatial Pyramid pooling, Mixed pooling, 6
Batch normalization.

Visualizing Neural Networks - Data oriented techniques, Gradient 7


based techniques, Inverting representation.
IV Classi cation of tra c signs - Dataset preparation, Training and
5
validation curves, Using ConvNets for tra c sign classi cation

Ensemble of ConvNets, Stability against Noise, Sliding Window within


ConvNets, Sparse Coding. 5

END SEMESTER EXAM (All Modules)

APJ Abdul Kalam Technological Universit


Page 16 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
fi
fi
ffi
y

fi
fi
ffi
fi
fi
fi
References:


1. Hamed Habibi Aghdam, Elnaz Jahani Heravi, “Guide to Convolutional Neural Networks : A
practical application to Tra c Sign Detection and Classi cation”, Springer, 2017

2. Ethem Alpaydin, "Introduction to Machine Learning”, MIT Press, Prentice Hall of


India,2005.

3. Tom Mitchell, “Machine Learning”, McGraw Hill, 3rd Edition,1997.

4. Christopher M. Bishop, “Pattern Recognition and Machine Learning”, Springer (2006).

5. Kevin P. Murphy, “Machine Learning: A Probabilistic Perspective”, The MIT Press, 2012

APJ Abdul Kalam Technological Universit


Page 17 of 73 Cluster 6 | M. Tech Programme in AI & DS
ffi
y

fi
Course Year of
Course Name L-T-P : C
Code Introduction

06DS6351 Soft Computing 3-0-0 : 3 2020

Course Objective:


1. To familiarise the salient approaches in soft computing, based on arti cial neural networks,
fuzzy logic, and genetic algorithms

2. To introduce applications of soft computing in di erent research areas in Computer Science/


Information Technology
SYLLABUS:
Basic concepts, learning models, ANN architectures, Fuzzy sets and logic, Crisp vs Fuzzy logic,
Genetic Algorithms, Hybrid systems.
Course Outcomes:


Students should be able to

1. Learn about soft computing techniques and their applications

2. Analyse various neural network architectures



3. Understand perceptrons and counter propagation networks.

4. De ne the fuzzy systems



5. Analyse the genetic algorithms and their applications.
Module Course Content Hours
I Basic concept of Soft Computing; Basic concept of neural networks,
Mathematical model, Properties of neural network, Typical
5
architectures: single layer, multilayer, competitive layer

Di erent learning methods: Supervised, Unsupervised & reinforced;


Common activation functions; Feed forward, Feedback & recurrent 5
N.N; Application of N.N
INTERNAL TEST 1 (Module I)

II Architecture, Algorithm & Application of McCulloh-Pitts, Hebb Net,


Perceptron (with limitations & Perceptron learning rule Convergence
11
theorem), Back propagation Neural Network, ADALINE, MADALINE,
Discrete Hop eld net, Bidirectional Associative Memory, Maxnet
INTERNAL TEST 2 (Module II)

III Fuzzy Sets & Logic : Fuzzy versus Crisp; Fuzzy sets—membership
function, linguistic variable, basic operators, properties; Fuzzy relations
6
—Cartesian product, Operations on relations;

Crisp logic—Laws of propositional logic, Inference; Predicate logic—


5
Interpretations, Inference; Fuzzy logic—Quanti ers, Inference;
Defuzzi cation methods.

APJ Abdul Kalam Technological Universit


Page 18 of 73 Cluster 6 | M. Tech Programme in AI & DS
ff
fi
fi
fi
y

ff
fi
fi
IV Genetic Algorithm (GA) Basic concept; role of GA in optimization,
Fitness function, Selection of initial population, Cross over(di erent 5
types), Mutation, Inversion, Deletion, Constraints Handling;
Evolutionary Computation.

Hybrid Systems : GA based BPNN(Weight determination); Neuro


Fuzzy Systems—Fuzzy BPNN--fuzzy Neuron, architecture, learning; 5
Fuzzy Logic controlled G.A.
END SEMESTER EXAM (All Modules)
References:


1. Neural Networks- A Comprehensive foundation, Simon Haykin, 2nd Ed; Pearson

2. Principles of Softcomputing, S.N. Sivanandam, S.N.Deepa, Wiley India.

3. Neural Networks, Fuzzy Logic & Genetic Algorithms – Synthesis & applications, T.S.
Rajasekaran & G.A. Vijaylakshmi Pai, PHI

4. Genetic Algorithm & fuzzy Logic Systems - Sanchez, Takanori, Zadeh; World Scienti c

5. Genetic Algorithm, Goldberg David E.; Pearson

6. Fuzzy Set Theory & Its Applications, Zimmermann H. J, Allied Publishers Ltd, 1991.

7. Neuro-Fuzzy and Soft Computing, A Computational Approach to Learning and Machine


Intelligence, Jyh-Shing Roger Jang, Chuen-Tsai Sun, Eiji Mizutani, Prentice-Hall of India Pvt.
Ltd., 2004. ISBN:978-0-13261-066-7

8. Fuzzy Logic with Engineering Applications (3rd Edn.), Timothy J. Ross, Willey, 2010.

APJ Abdul Kalam Technological Universit


Page 19 of 73 Cluster 6 | M. Tech Programme in AI & DS
y

ff
fi
Course Year of
Course Name L-T-P : C
Code Introduction

06DS6451 Computer Vision 3-0-0 : 3 2020

Course Objectives:

1. Introduce the standard computer vision problems and identify the solution methodologies.

2. To introduce concepts of Linear discriminant based and tree based classi ers

SYLLABUS:
Image formation and modelling, A ne structures, Bayesian decision theory, Linear discriminants
and Trees, Pattern recognition.

Course Outcomes:


Students should be able to

1. Understand and implement the algorithms for 3D reconstruction from various cues. 

2. Understand and implement the various segmentation, pattern analysis, objection 

detection/recognition methods.
Module Course Content Hours
I Image formation and Image model- Components of a vision system-
Cameras- camera model and camera calibration- Radiometry- Light in
8
space- Light in surface - Sources, shadows and shading, Multiple
images-The Geometry of multiple views- Stereopsis
INTERNAL TEST 1 (Module I)
II A ne structure from motion- Elements of A ne Geometry, A ne
structure and motion from two images- A ne structure and motion 6
from multiple images- From A ne to Euclidean images.

High level vision- Geometric methods- Model based vision- Obtaining


hypothesis by pose consistency, pose clustering and using Invariants, 6
Veri cation.
INTERNAL TEST 2 (Module II)

III Bayesian Decision Theory- Minimum error rate classi cation


Classi ers, discriminant functions, decision surfaces- The normal 7
density and discriminant-functions for the Normal density.
IV Linear discriminant based classi ers and tree classi ers - Linear
discriminant function based classi ers- Perceptron- Minimum Mean 9
Squared Error (MME) method, Support Vector machine, Decision Trees:
CART, ID3.

Recent Advances in Pattern Recognition - Neural network structures 6


for pattern recognition, Pattern classi cation using Genetic Algorithms.
END SEMESTER EXAM (All Modules)

APJ Abdul Kalam Technological Universit


Page 20 of 73 Cluster 6 | M. Tech Programme in AI & DS
ffi
fi
fi
y

ffi
ffi
fi
fi
fi
ffi
ffi
fi
fi
ffi
fi
References:
1. C. M. Bishop, Pattern Recognition and Machine Learning, Springer, 2006.

2. R. O. Duda, P. E. Hart and D. G. Stork, Pattern Classi cation, John Wiley, 2001.

3. Richard Hartley and Andrew Zisserman, Multiple View Geometry in Computer



Vision, Second Edition, Cambridge University Press, 2004.

4. S. Theodoridis and K. Koutroumbas, Pattern Recognition, 4th Ed., Academic Press, 2009.

APJ Abdul Kalam Technological Universit


Page 21 of 73 Cluster 6 | M. Tech Programme in AI & DS
y

fi
Course Year of
Course Name L-T-P : C
Code Introduction

06DS6061 Research Methodology 0-2-0 : 2 2020

Course Objective:


1. To prepare the student to do the M. Tech project works with a research bias.

2. To formulate a viable research question.

3. To develop skill in the critical analysis of research articles and reports.

4. To analyze the bene ts and drawbacks of di erent methodologies.

5. To understand how to write a technical paper based on research ndings.


SYLLABUS:
Introduction to Research Methodology - Types of research - Ethical issues - Copy right - royalty
- Intellectual property rights and patent law - Copyleft - Open access - Analysis of sample
research papers to understand various aspects of research methodology : De ning and
formulating the research problem - Literature review - Development of working hypothesis -
Research design and methods - Data Collection and analysis - Technical writing - Project work
on a simple research problem
Course Outcomes:


Students should be able to

1. Understand research concepts in terms of identifying the research problem

2. Propose possible solutions based on research

3. Write a technical paper based on the ndings

4. Get a good exposure to a domain of interest

5. Get a good domain and experience to pursue future research activities
Module Course Content Hours
I Introduction to Research Methodology: Motivation towards research
- Types of research: Find examples from literature.

Professional ethics in research - Ethical issues-ethical committees.


Copy right - royalty - Intellectual property rights and patent law - Copy
6
left - Open access - Reproduction of published material - Plagiarism -
Citation and acknowledgement.

Impact factor. Identifying major conferences and important journals in


the concerned area. Collection of at least 4 papers in the area.
INTERNAL TEST 1 (Module I)

II De ning and formulating the research problem - Literature Survey -


Analyze the chosen papers and understand how the authors have
undertaken literature review, identi ed the research gaps, arrived at
their objectives, formulated their problem and developed a hypothesis. 8

INTERNAL TEST 2 (Module II)

APJ Abdul Kalam Technological Universit


Page 22 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
fi
y

fi
fi
ff
fi
fi
III Research design and methods: Analyze the chosen papers to
understand formulation of research methods and analytical and
experimental methods used. Study of how di erent it is from previous
works.

7
Data Collection and analysis. Analyze the chosen papers and study
the methods of data collection used. - Data Processing and Analysis
strategies used– Study the tools used for analyzing the data.
IV Technical writing - Structure and components, contents of a typical
technical paper, di erence between abstract and conclusion, layout,
illustrations and tables, bibliography, referencing and footnotes-use of
tools like Latex.

7
Identi cation of a simple research problem – Literature survey-
Research design- Methodology –paper writing based on a hypothetical
result.

References:

1. C. R. Kothari, Research Methodology, New Age International, 2004

2. Panneerselvam, Research Methodology, Prentice Hall of India, New Delhi, 2012.

3. J. W. Bames, Statistical Analysis for Engineers and Scientists, Tata McGraw-Hill, New York.

4. Donald Cooper, Business Research Methods, Tata McGraw-Hill, New Delhi.

5. Leedy P. D., Practical Research: Planning and Design, McMillan Publishing Co.

6. Day R. A., How to Write and Publish a Scienti c Paper, Cambridge University Press, 1989.

7. Manna, Chakraborti, Values and Ethics in Business Profession, Prentice Hall of India, New 

Delhi, 2012.

8. Sople, Managing Intellectual Property: The Strategic Imperative, Prentice Hall ofIndia, New
Delhi, 2012.

9. Vinod Chandra S. S., Anand H. S. Research Methodology, Pearson Education, ISBN:


978-93-528-6351-8, 2017

APJ Abdul Kalam Technological Universit


Page 23 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
ff
y

fi
ff
Course Year of
Course Name L-T-P : C
Code Introduction

06DS6071 Seminar I 0-0-2 : 2 2020

Course Objectives:


1. To introduce the students to research, make them understand research papers and prepare
presentation material 

2. To understand cutting edge technology in the chosen area

3. To improve oral communication skills through presentation

4. To prepare original technical write up on the presentation

Course Outcomes:

After completion of course, students will be able to:



1. Develop skills in doing literature survey, technical presentation and report preparation

2. Improve the pro ciency in English



3. Improve presentation skills

4. Improve analytical and reasoning ability

5. Improve technical writing skills

Syllabus:

The aim of this course is to introduce the student to research, and to acquaint him with the process of
presenting his work through seminars and technical reports. Students have to register for the seminar and
select a topic in consultation with any faculty member o ering courses for the programme. The student is
expected to do an extensive literature survey and analysis in an area related to computer science (other
than the area of specialisation). The study should preferably result in design ideas, designs, algorithms,
and theoretical contributions in the form of theorems and proofs, new methods of proof, new techniques
or heuristics with analytical studies, implementations and analysis of results.

The presentation shall be of 30 minutes duration and a committee with the Head of the Department as the
chairman and two faculty members from the department as members shall evaluate the seminar based on
the coverage of the topic, presentation and ability to answer the questions put forward by the committee.

Students shall individually prepare and submit a seminar report based on experimental study / industrial
training on the corresponding topic, in the prescribed format given by the Department. The reference shall
include standard journals (ACM/IEEE), conference proceedings and equivalent documents, reputed
magazines and textbooks, technical reports and web based material, approved by the supervisor. The
references shall be incorporated in the report following IEEE standards re ecting the state-of-the-art in
the topic selected.

APJ Abdul Kalam Technological Universit


Page 24 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
y

ff
fl
Course Year of
Course Name L-T-P : C
Code Introduction

06DS6081 Machine Intelligence Lab 0-0-3 : 1 2020

Course Objectives:

1. Implement basic algorithm in AI

2. Make use of Data sets in implementing the machine learning algorithms

3. Implement the machine learning concepts and algorithms in any suitable language of choice
Course Outcomes:


Students should be able to

1. Apply AI algorithms to solve real world problems

2. Understand the implementation procedures for the machine learning algorithms

3. Design Java/Python programs for various Learning algorithms

4. Apply appropriate data sets to the Machine Learning algorithms

5. Identify and apply Machine Learning algorithms to solve real world problems 


Sl No List of Experiments

1 Introduction to Python-based notebook environments

2 Implement A* algorithm for one the following problems: i) 8 puzzle ii) Missionaries and
Cannibals 
3 Implement and test hill climbing based search algorithms to solve Travelling Salesman
Problem
4 Solve and implement map coloring problem by backtracking and constraint
propagation
5 Solve and implement the game of tic-tac-toe using mini-max

6 Write a program to demonstrate the working of the decision tree based ID3 algorithm.
Use an appropriate data set for building the decision tree and use it to classify a new
sample
7 Build an Arti cial Neural Network by implementing the Backpropagation algorithm and
test the same using appropriate data sets
8 Write a program to implement the naive Bayesian classi er for a sample training data
set stored as a .CSV le. Compute the accuracy of the classi er, considering few test
data sets. Calculate the accuracy, precision, and recall for your data set
9 Write a program to construct a Bayesian network considering medical data. Use this
model to demonstrate the diagnosis of heart patients using any standard Heart Disease
Data Set
10 Apply EM algorithm to cluster a set of data stored in a .CSV le. Use the same data set
for clustering using k-Means algorithm. Compare the results of these two algorithms
and comment on the quality of clustering
11 Write a program to implement k-Nearest Neighbour algorithm to classify the iris data
set. Print both correct and wrong predictions

APJ Abdul Kalam Technological Universit


Page 25 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
fi
y

fi
fi
fi
12 Implement the non-parametric Locally Weighted Regression algorithm in order to t
data points. Select appropriate data set for your experiment and draw the
corresponding graphs
13 Write a program to implement 5-fold cross validation on a given dataset. Compare the
accuracy, precision, recall, and F-score for your data set for di erent folds
Ten experiments to complete mandatory

APJ Abdul Kalam Technological Universit


Page 26 of 73 Cluster 6 | M. Tech Programme in AI & DS
y

ff
fi
APJ Abdul Kalam Technological University
Master of Technology – Course Plan

SEMESTER II

M. Tech Programme in
Arti cial Intelligence and Data Scienc

APJ Abdul Kalam Technological Universit


Page 27 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
y



e

Course Year of
Course Name L-T-P : C
Code Introduction

06DS6012 Big Data Analytics 4-0-0 : 4 2020

Course Objectives:

1. To familiarise the Big Data Platform and its use cases

2. To provide an overview of Apache Hadoop

3. To provide HDFS Concepts and Interfacing with HDFS

4. To familiarise Map Reduce analytics using Hadoop and related tools like Pig, Hive etc.
SYLLABUS:
Introduction, Nature of data, Mining data streams, Estimations, Predictive analysis,
Visualizations, Hadoop, HDFS, MapReduce, Case Study.
Course Outcomes:


Students should be able to

1. Describe big data and use cases from selected business domains

2. Explain the components of Hadoop and Hadoop Eco-System

3. Install, con gure, and run Hadoop and HDFS

4. Perform map-reduce analytics using Hadoop

5. Use Hadoop related tools such as HBase, Pig, and Hive for big data analytics
Module Course Content Hours
I Introduction to big data: Introduction to Big Data Platform,
Challenges of Conventional Systems - Intelligent data analysis, Nature 12
of Data - Analytic Processes and Tools - Analysis vs Reporting.
INTERNAL TEST 1 (Module I)

II Mining data streams: Introduction To Streams Concepts, Stream


Data Model and Architecture - Stream Computing - Sampling Data in a 7
Stream, Filtering Streams, Counting Distinct Elements in a Stream

Estimating Moments, Counting Oneness in a Window, Decaying


Window - Real time Analytics Platform (RTAP) Applications - Case 7
Studies - Real Time Sentiment Analysis- Stock Market Predictions.
INTERNAL TEST 2 (Module II)

III Components of Hadoop - Analysing the Data with Hadoop- Scaling


Out- Hadoop Streaming- Design of HDFS-Java interfaces to HDFS
6
Basics

Developing a Map Reduce Application-Anatomy of a Map Reduce


Job, Scheduling-Shuffle and Sort - Task execution. Case Study: IBM 6
InfoSphere BigInsights and Streams.

APJ Abdul Kalam Technological Universit


Page 28 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
y

IV Introduction to HBase, Filesystems for HBase, Client API - The


Basics, Hbase clients – REST, Shell Commands, Map Reduce
Integration

7
Introduction to Pig, Grunt, pig data model, Pig Latin, Advanced pig
latin, developing and testing Pig Latin scripts, Map Reduce Integration

7
Hive, data types and le formats, HiveQL data de nition, HiveQL data
manipulation, HiveQL queries, HiveQL views, HiveQL Indexes,
functions.
END SEMESTER EXAM (All Modules)
References:


1. Michael Minelli, Michelle Chambers, and AmbigaDhiraj, "Big Data, Big Analytics:
Emerging Business Intelligence and Analytic Trends for Today's Businesses", Wiley,2013.
2. P. J. Sadalage and M. Fowler, "NoSQL Distilled: A Brief Guide to the Emerging World of
Polyglot Persistence", Addison-Wesley Professional,2012.
3. Tom White, "Hadoop: The Definitive Guide", Third Edition, O'Reilley,2012. Analytics for
Enterprise Class Hadoop and Streaming Data”, McGrawHill Publishing, 2012.
4. Eric Sammer, "Hadoop Operations", O'Reilley,2012.
5. E. Capriolo, D. Wampler, and J. Rutherglen, "Programming Hive", O'Reilley,2012.
6. Lars George, "HBase: The Definitive Guide", O'Reilley,2011.
7. Alan Gates, "Programming Pig", O’Reilley,2011.
8. Chris Eaton, Dirk De Roos, Tom Deutsch, George Lapis, Paul Zikopoulos,“Understanding
Big Data: Analytics for Enterprise Class Hadoop and Streaming Data”, McGrawHill
Publishing, 2012.
9. Anand Rajaraman and Jeffrey D. Ullman,”Mining of Massive Datasets”, Cambridge
University Press, 2012.
10. Arshdeep Bahga, Vijay Madisetti, “Big Data Science & Analytics: A Hands - On
Approach",VPT, 2016

APJ Abdul Kalam Technological Universit


Page 29 of 73 Cluster 6 | M. Tech Programme in AI & DS

fi

fi

Course Year of
Course Name L-T-P : C
Code Introduction

06DS6022 Deep Learning & Arti cial Neural Network 3-0-0 : 3 2020

Course Objectives:


1. To familiarise and master the tools of Artificial Intelligence


2. To explore in depth deep neural architectures for learning and inference
3. To evaluate the performance of neural architectures in comparison to other machine learning
methods
SYLLABUS:
Neural networks, Perceptrons, Backpropagation, Deep Networks, Deep Reinforcement learning,
RNN, LSTM, Deep Unsupervised learning, Adversarial learning, Deep Generative Models.
Course Outcomes:


Students should be able to
1. Understand basic Neural Network architectures
2. Apply fundamental principles, theory and approaches for learning with deep neural networks
3. Analyse main variants of deep learning and their typical applications
4. Analyse how deep learning fits within the context of other Machine Learning approaches
Module Course Content Hours

I Neural networks- Perceptrons, sigmoid units; Learning in neural


networks - output vs hidden layers; linear vs nonlinear networks; linear 6
models (regression) - LMS algorithm.

Perceptrons classi cation - limitations of linear nets and perceptrons


- multi-Layer Perceptrons (MLP)- activation functions - linear, softmax, 6
tanh, ReLU; error functions - feed-forward networks.
INTERNAL TEST 1 (Module I)

II Backpropagation - recursive chain rule - Learning weights of a logistic 7


output neuron - loss functions - learning via gradient descent -
optimization momentum method; Adaptive learning rates RmsProp -
mini-batch gradient descent - bias-variance trade o , regularization -
over tting - inductive bias regularization - drop out - generalization.

Deep neural networks - convolutional nets case studies using Keras/


2
Tensor ow.
INTERNAL TEST 2 (Module II)

III Introduction to deep reinforcement learning - neural nets for


sequences - Recurrent Nets, LSTM
5

Introduction to Deep unsupervised learning autoencoders - PCA to


autoencoders - Deep Generative Models - Generative Models and
Variational Inference - Autoregressive Models and Invertible 6
Transformations

APJ Abdul Kalam Technological Universit


Page 30 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
fl
fi
fi

ff

IV Adversarial Learning - Unifying Variational Autoencoders and 5


Generative Adversarial Networks - Adverserial Autoencoders -
Evaluation of Generative Models

Geometry of Deep Generative Models - Application - Model based


Reinforcement Learning. 5

END SEMESTER EXAM (All Modules)

References:

1. Ian Goodfellow, Yoshua Bengio, Aaron Courville. Deep Learning, Second edition, MITPress,
2016
2. Duda R.O., Hart P.E., Stork D.G., Pattern Classification, Second edition, Wiley - Interscience,
2001
3. Theodoridis, S., Koutroumbas, K. Pattern Recognition, Fourth edition, Academic Press,2008
4. Russell S., Norvig N., Artificial Intelligence: A Modern Approach, Prentice Hall Series in
Artificial Intelligence,2003
5. Bishop C.M. Neural Networks for Pattern Recognition, Oxford University Press,1995
6. HastieT.,Tibshirani R.and FriedmanJ .,The Elements of Statistical Learning, Springer,2001
7. Koller D.and Friedman N. Probabilistic Graphical Models, MITPress, 2009

APJ Abdul Kalam Technological Universit


Page 31 of 73 Cluster 6 | M. Tech Programme in AI & DS

Course Year of
Course Name L-T-P : C
Code Introduction

06DS6032 Genetic Algorithms 3-0-0 : 3 2020

Course Objectives:


1. To familiarise the basic background of genetic algorithm

2. To explain how NP problems can be tried solving using genetic algorithm strategies
SYLLABUS:
Evolutionary computation, Genetic Algorithms, Encoding, Steady state algorithms, Genetic
programming, GA in engineering, GA in optimization, GA in scienti c models and theoretical
foundations, GBML.
Course Outcomes:


Students should be able to
1. Explain the of the principles underlying Evolutionary Computation in general and Genetic
Algorithms in particular.

2. Apply Evolutionary Computation Methods to nd solutions to complex problems 

3. Summarise current research in Genetic Algorithms and Evolutionary Computing 


Module Course Content Hours
I A brief history of evolutionary computation-biological terminology-
search space -encoding, reproduction-elements of genetic algorithm- 8
genetic modeling- Comparison of GA and traditional search methods.
INTERNAL TEST 1 (Module I)

II Steady state algorithm - tness scaling - inversion.


6

Genetic programming - Genetic Algorithm in problem solving,


Implementing GP. 6

INTERNAL TEST 2 (Module II)

III Genetic Algorithm in engineering and optimization- natural 7


evolution - simulated annealing and Tabu search.

Genetic Algorithm in scienti c models and theoretical foundations


8
- computer implementation - low level operator and knowledge based
techniques in Genetic Algorithm.
IV Applications of Genetic based machine learning-Genetic Algorithm
and parallel processors, constraint optimization, uses of GA in solving 7
NP hard problems, multilevel optimization, real life problem.
END SEMESTER EXAM (All Modules)

APJ Abdul Kalam Technological Universit


Page 32 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
y

fi
fi
fi
References:

1. Melanie Mitchell, “An introduction to Genetic Algorithm”, Prentice-Hall of India, New Delhi,
Edition: 2004.

2. David.E.Golberg, “Genetic algorithms in search, optimization and machine learning”,


Addition-Wesley-1999.

3. S.Rajasekaran G.A Vijayalakshmi Pai, “Neural Networks, Fuzzy logic and Genetic
Algorithms Synthesis and Applications”, Prentice Hall of India, New Delhi-2003.

4. Nils.J.Nilsson, “Arti cial Intelligence- A new synthesis”, Original edition-1999.

APJ Abdul Kalam Technological Universit


Page 33 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
y

Course Year of
Course Name L-T-P : C
Code Introduction

06DS6142 R for Data Science 3-0-0 : 3 2020

Course Objectives:


1. Introduction to data science life cycle

2. In depth knowledge of most popular machine learning techniques 



3. Supervised and unsupervised learning techniques

4. Real life case studies and simulated projects to sharpen your skill sets

5. Assistance in creating a portfolio which will allow you to showcase your newly acquired skills
SYLLABUS:
R environment, Data loading and organization, factors, Data exploration and cleaning, Matrices,
Machine Learning models, Statistical models, Documentation and plotting.

Course Outcomes:


Students should be able to

1. Analyse data and nd relative patterns to predict outcomes

2. Analyse continuous data in varying scenarios

3. Perform Con rmatory Data analysis

4. Demonstrate expert knowledge in outcome predictions


Module Course Content Hours
I Introduction, Reading and getting data into R, Vectors and
assignment, Logical and Index vectors, Generating regular sequences,
10
Missing values, Ordered and Unordered Factors, The function tapply()
and ragged arrays, Ordered factors, Reading data from les.
INTERNAL TEST 1 (Module I)

II Exploring and cleaning data for analysis, Data organization, Arrays


and Matrices, Basics of Arrays in R, Matrix operations, Advanced 10
Matrix operations, Additional Matrix facilities, Lists and Data frames.
INTERNAL TEST 2 (Module II)

III Mapping models to Machine Learning, Evaluating and Validating


models, Probability distributions in R, Statistical models in R, Building
11
linear models, Generalized linear models, Nonlinear least squares and
maximum likelihood models.
IV Documentation, Graphical analysis, plot() function, Displaying
multivariate data, Using graphics parameters, Matrix plots, Exporting 11
graphs, ggplot package.
END SEMESTER EXAM (All Modules)

APJ Abdul Kalam Technological Universit


Page 34 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
fi
y

fi
References:


1. Jure Leskovec, Anand Rajaraman, Je rey D. Ullman, "Mining of Massive Datasets".


Cambridge University Press, 2014

2. Nathan Yau, "Visualize This: The Flowing Data Guide to Design, Visualization and
Statistics", Wiley, 2011

3. Nina Zumel, John Mount "Practical Data Science with R··. Manning Publications. 2014

4. Tony Ojeda, Sean Patrick Murphy, Benjarnin Bengfort. Abhijit Dasgupta. "Practical Data
Science Cookbook", Packt Publishing Limited, 2014

5. W. N. Venables. D. M. Smith and the R Core Team, "An Introduction to R", 2013

APJ Abdul Kalam Technological Universit


Page 35 of 73 Cluster 6 | M. Tech Programme in AI & DS
y

ff
Course Year of
Course Name L-T-P : C
Code Introduction

06DS6242 Data Analytics & Scalable Algorithms 3-0-0 : 3 2020

Course Objectives:


1. Familiarise how data a ects the performance of analysis algorithms and their scalability.

2. Identify the role of attributes in parallelisation of analytics algorithms.

3. Explore the use of pre-trained models in data analytics.

SYLLABUS:
Data and Relations, Correlation, Analytical models, Clustering, ReLU, Augmentation,
Convolution and Pooling, Parallelization, Hashing, Batch processing, Available networks.
Course Outcomes:


Students should be able to

1. Identify data errors and dependencies among attributes by modelling them as sets & relations

2. Apply regression, classi cation, and clustering models on a given dataset

3. Analyse data and processes for opportunities on parallelisation

4. Apply data on pertained networks and perform classi cation


Module Course Content Hours
I Data and Relations - Data scales, Set and Matrix representations,
Relations, Similarity and dissimilarity measures, Sequence relations. 6
Data preprocessing - Error types, error handling, ltering,
transformation, merging. Data visualisation.

Correlation - Linear, Causality, Chi-Square tests. Regression - Linear


regression, Robust regression, RBF networks, Cross validation and 6
feature selection.
INTERNAL TEST 1 (Module I)

II Finite state machines, Recurrent models, Autoregressive models, Naive


Bayes classi er, LDA, SVM, Learning Vector Quantization.
5

Cluster partitions, Sequential clustering, Prototype based clustering,


Fuzzy clustering, Relational clustering, Cluster tendency assessment, 5
Cluster validity, Self organising map.
INTERNAL TEST 2 (Module II)

III ReLU nonlinearity, Data Augmentation, MLP Convolutional Layer,


Global Average Pooling, Dimensionality Reduction, Cascading, CNN- 6
based feature extraction.

Scalability through parallelization - Data parallelization, Process


parallelization, Scaling using feature engineering, Feature reduction 5
through spatial transforms.
IV Use of hashing, Multiple feature hashing, Multimodal fusion for 6
classi cation, Batch processing frameworks.

Case Studies : AlexNet, VGG, GoogLeNet, ResNet 3


END SEMESTER EXAM (All Modules)

APJ Abdul Kalam Technological Universit


Page 36 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
fi
ff
fi
y

fi
fi
References:


1. Thomas A. Runkler, “Data Analytics - Models and Algorithms for Intelligent Data Analysis”,
Springer 2012.

2. Stefanos Vrochidis, Benoit Huet, Edward Chang, Ioannis Kompatsiaris, “Big Data Analysis
for Large-Scale Multimedia Search”, Wiley 2019.

3. J. O. Moreira, Andre Carvalho, Tomas Horvath, “A General Introduction to Data Analytics”,


Wiley 2019.

APJ Abdul Kalam Technological Universit


Page 37 of 73 Cluster 6 | M. Tech Programme in AI & DS
y

Course Year of
Course Name L-T-P : C
Code Introduction

06DS6342 Scalable Systems for Data Science 3-0-0 : 3 2020

Course Objectives:


1. Teach the fundamental systems aspects of designing using Big Data platforms

2. Explore distributed program models and abstractions


SYLLABUS:
Distributed systems, Distributed File System, Network protocols, Hadoop, Parallel Data Mining
agents, MapReduce, Grid technology.
Course Outcomes:


Students should be able to

1. Distinguish distributed programming models for Big Data like Map Reduce, Stream
processing and Graph processing.

2. Design and develop applications on Big Data platforms and their optimisations on commodity
clusters and Clouds.

3. Scale data science algorithms and analytics using Big Data platforms.
Module Course Content Hours
I Introduction to Distributed Systems, evolution, characteristics,
design issues, user requirements, Distributed computing models-
8
workstation model, workstation-server model, processor–pool model.
Protocols for distributed systems -VMTP and FLIP.
INTERNAL TEST 1 (Module I)

II Distributed le system: Components of DFS, design issues,


interfaces, implementation, File Caching and Replication. Sun Network 6
File System – architecture and implementation, Google File System.
Naming- Namespace and contexts and name resolution.

Network Protocols, Naming, RPC, RMI, Web Services, CORBA, A


message passing model for Inter-Process Communication,
Coordination algorithms, Leader Election, Bully Algorithm, Maxima 7
Finding on a Ring
INTERNAL TEST 2 (Module II)

III Hadoop Architecture - Clusters, HDFS, YARN, Basic le system


operations in HDFS, File permissions in HDFS, Functional
10
Programming Model of MapReduce, Job Chaining, Submitting
MapReduce job to YARN
IV Parallel Data Mining Agents, Parallel Data Access, Parallel Data
Analysis, Parallel GA in Big Data Analysis, Evolutionary Algorithm
Based Techniques to Handle Big Data, Statistical and Evolutionary
11
Feature Selection Techniques Parallelized Using MapReduce
Programming Model, The Role of Grid Technologies: A Next Level
Combat with Big Data
END SEMESTER EXAM (All Modules)

APJ Abdul Kalam Technological Universit


Page 38 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
y

fi
References:

1. Sunita Mahajan, Seema shah, Distributed Computing ,Oxford University Press, rst edition,
2010

2. George Coulouris, Jean Dellimore and Tim Kindberg, Distributed Systems – Concepts and
designing, Pearson Education Asia, Fifth Edition 2006, New Delhi.

3. Pradeep. K, Sinha, Distributed Operating Systems ,PHI Edition, rst Edition,1997.

4. Andrew S Tenenbaum, Distributed Operating Systems, Pearson Education Asia

5. Distributed Systems An Algorithmic Approach, Sukumar Ghosh, CRC Press, 2007

6. Techniques and Environments for Big Data Analysis : Parallel, Cloud, and Grid Computing,
Studies in Big Data Vol 17, 2016

7. Web based Parallel / Distributed Medical Data Mining Using Software Agents - Hillol
Kargupta, Brian Sta ord, Ilker Hamzaoglu, Los Alamos National Labs, 1997

8. Data Analytics with Hadoop: An Introduction for Data Scientists

APJ Abdul Kalam Technological Universit


Page 39 of 73 Cluster 6 | M. Tech Programme in AI & DS
ff
y

fi
fi
Course Year of
Course Name L-T-P : C
Code Introduction

06DS6442 Knowledge Engineering and Data Science 3-0-0 : 3 2020

Course Objectives:


1. To explore the practical application of intelligent technologies into the di erent domains

2. To give students insight and experience in key issues of data and knowledge processing
SYLLABUS:
Formalisms, Items and Objects, Schema and normalization, Analysis models, Evidence and
knowledge, Analysis and synthesis, Ontology.
Course Outcomes:


Students should be able to

1. Understand and describe the concepts central to the creation of knowledge bases and expert
systems.

2. Conduct an in-depth examination of an existing expert system with an emphasis on basic


methods of creating a knowledge base.

3. Demonstrate pro ciency with statistical analysis of data.

4. Build and assess data-based models.


Module Course Content Hours
I Formalisms - Logic as a programming language, Logic as a
knowledge language, Logic as a database language, lambda calculus, 5
Data, information and knowledge, Knowledge based systems.

Items and Objects - uni ed representation, structure of data,


information, and knowledge items, structure of object, data, 5
information, and knowledge objects. Algebra of objects.
INTERNAL TEST 1 (Module I)

II Schema and normalization - r-schema and i-schema, o-schema, t-


schema, Classical normal forms.
5

Analysis - conceptual view of objects, c-coupling map, constraints.


Functional model - functional view, f-coupling map, constraints. Layout 6
- internal view, i-coupling.
INTERNAL TEST 2 (Module II)

III Evidence and Knowledge, Abductive Reasoning, Probabilistic


Reasoning, Belief functions, Baconian and Fuzzy probability, Evidence
based reasoning. Ontology of problem solving tasks, Building 6
knowledge based agents. Agent Design and Development using
Learning Technology.

Problem solving through analysis and synthesis, Inquiry driven


5
analysis and synthesis for Evidence-based reasoning, Believability
assessment.

APJ Abdul Kalam Technological Universit


Page 40 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
y

fi
ff
IV Ontology Design and Development, Reasoning with ontologies and
rules - Reduction and synthesis rules, Rule and ontology matching,
Partially learned knowledge, Reasoning with partially learned
10
knowledge. Generalization and specialization for knowledge based
agents, Rule learning - Analogy-based generalization, Hypothesis
learning.
END SEMESTER EXAM (All Modules)
References:


1. John Debenham, “Knowledge Engineering - Unifying Knowledge Base and Database


Design”, Springer 2007.

2. Jude Hemnath, Madhulika Bhatia, Oana Geman, “Data Visualization and Knowledge
Engineering”, Springer 2020.

3. Gheorghe Tecuci, Dorin Marcu, Mihai Boicu, David A. Schum, “Knowledge Engineering -
Building Cognitive Assistants for Evidence Based Reasoning”, Cambridge University Press,
2016.

APJ Abdul Kalam Technological Universit


Page 41 of 73 Cluster 6 | M. Tech Programme in AI & DS
y

Course Year of
Course Name L-T-P : C
Code Introduction

06DS6152 Big Data for Internet of Things 3-0-0 : 3 2020

Course Objectives:


1. Familiarise the scale of data generated from Internet of Things

2. Identify semantic relationships between di erent pieces of data generated from IoT

3. Explore the possibility of integrating cloud infrastructure to analyse data from IoT

SYLLABUS:
Big Data platforms for IoT, Challenges in IoT environment, Spatial dimensions of data, Fog
Computing, Automation with Web, Data and Analysis in cloud
Course Outcomes:


Students should be able to

1. Understand aspects of volume, velocity, variety, veracity, and variability in IoT-based data

2. Understand fog computing as an infrastructure for load balancing in IoT-based data analysis

3. Apply concepts in IoT and data analysis to design smart systems

4. Apply cloud computing technologies to manage high-scale data from IoT and social networks

Module Course Content Hours


I Big Data Platforms for the Internet of Things: network protocol- data
dissemination – current state of art- Improving Data and Service
Interoperability with Structure, Compliance, Conformance and
Context Awareness: interoperability problem in the IoT context- Big
Data Management Systems for the Exploitation of Pervasive 11
Environments - Big Data challenges and requirements coming from
di erent Smart City applications. Adaptive Pipelined Neural Network
Structure in Self- aware Internet of Things: self-healing systems- Role
of adaptive neural network.
INTERNAL TEST 1 (Module I)

II Spatial Dimensions of Big Data: Application of Geographical


Concepts and Spatial Technology to the Internet of Things- Applying
spatial relationships, functions, and models. Fog Computing: A
10
Platform for Internet of Things and Analytics: a massively distributed
number of sources - Big Data Metadata Management in Smart Grids:
semantic inconsistencies – role of metadata.
INTERNAL TEST 2 (Module II)

III Toward Web Enhanced Building Automation Systems: heterogeneity


between existing installations and native IP devices - loosely-coupled
Web protocol stack –energy saving in smart building- Intelligent
Transportation Systems and Wireless Access in Vehicular Environment
11
Technology for Developing Smart Cities: advantages and
achievements- Emerging Technologies in Health Information Systems:
Genomics Driven Wellness Tracking and Management System (GO-
WELL) – predictive care – personalized medicine.

APJ Abdul Kalam Technological Universit


Page 42 of 73 Cluster 6 | M. Tech Programme in AI & DS
ff
y

ff
IV Data and Analytics in Cloud-Based M2M Systems - potential
stakeholders and their complex relationships to data and analytics
applications - Social Networking Analysis - Building a useful
10
understanding of a social network - Leveraging Social Media and IoT to
Bootstrap Smart Environments : lightweight Cyber Physical Social
Systems - citizen actuation.
END SEMESTER EXAM (All Modules)
References:


1. Stackowiak, R., Licht, A., Mantha, V., Nagode, L.,” Big Data and The Internet of Things
Enterprise Information Architecture for A New Age”, Apress, 2015.

2. Dr. John Bates , “Thingalytics - Smart Big Data Analytics for the Internet of Things”, john
Bates, 2015.

APJ Abdul Kalam Technological Universit


Page 43 of 73 Cluster 6 | M. Tech Programme in AI & DS
y

Course Year of
Course Name L-T-P : C
Code Introduction

06AD6252 Arti cial Intelligence and Robotics 3-0-0 : 3 2020

Course Objectives:


1. To familiarise the principles of reinforcement learning which is one of the key learning
techniques for robots

2. To familiarise uncertainty handling in robotics through probabilistic approaches 



3. To learn how measurements work for robots

SYLLABUS:
Overview of robotics, Dynamic programming, Approximate solutions, Recursive state
estimation, Filters, Measurement.
Course Outcomes:


Students should be able to

1. Learn the foundations of reinforcement learning for robotics 

2. Understand basic probabilistic principles behind Robotics intelligence 

3. Learn di erent measurement techniques for robotics 

4. Understand POMDP and its signi cance for robotics 

5. Implement principles of robotics intelligence for solving real world problems

Module Course Content Hours

I Overview: Robotics introduction, historical perspective on AI and 11


Robotics, Uncertainty in Robotics Reinforcement Learning: Basic
overview, examples, elements, Tabular Solution Methods - Multi-
armed bandits, Finite Markov decision process, Dynamic
programming (Policy Evaluation, Pol- icy Iteration, Value Iteration),
Monte Carlo Methods, Temporal-Di erence Learning (Q-learning,
SARSA)
INTERNAL TEST 1 (Module I)

II Approximate Solution Methods - On-policy Prediction with 10


Approximation, Value function approximation, Non-linear function
approximation, Reinforcement Learning in robotics
INTERNAL TEST 2 (Module II)

III Recursive state estimation: Robot Environment Interaction, Bayes 11


lters, Gaussian lters The Kalman lter, The Extended Kalman Filter,
The information lter, The particle lter Robot motion: Velocity Motion
Model, Odometry Motion Model, Motion and maps
IV Measurement: Beam Models of Range Finders, Likelihood Fields for 10
Range Finders, Correlation- Based Sensor Models, Feature-Based
Sensor Models, Overview of POMDP
END SEMESTER EXAM (All Modules)

APJ Abdul Kalam Technological Universit


Page 44 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
fi
ff
fi
fi
y

fi
fi
fi
ff
References:


1. Sebastian Thrun, Wolfram Burgard, Dieter Fox, Probabilistic Robotics, MIT Press, 2005 

2. Richard S. Sutton, Andrew G Barto, Reinforcement Learning: An Introduction, 2nd edition,
MIT Press, 2018

3. Jens Kober, Jan Peters, Learning Motor Skills: From Algorithms to Springer, 2014 

4. Francis X. Govers, Arti cial Intelligence for Robotics, Packt, 2018

APJ Abdul Kalam Technological Universit


Page 45 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
y

Course Year of
Course Name L-T-P : C
Code Introduction

06AD6352 Natural Language Processing 3-0-0 : 3 2020

Course Objectives:


1. To introduce the fundamental concepts and theory of Natural Language Processing (NLP) and
its practical applications

2. To explore Linguistic and statistical approaches to language processing in the three major
sub elds of NLP

SYLLABUS:
Introduction to NLP, N-grams, Neural networks in NLP, Vector semantics and embeddings,
Sentiment classi cation, POS tagging, Sequence processing with RNN.
Course Outcomes:


Students should be able to

1. Understand approaches to syntax and semantics in NLP

2. Understand approaches to generate, dialogue and summarisation within NLP

3. Understand current methods for statistical approaches to machine translation

4. Understand machine learning techniques used in NLP, including hidden Markov 

models and unsupervised methods
Module Course Content Hours
I Introduction – What is Natural Language Processing (NLP) - Syntax,
semantics, pragmatics, and ambiguity in NLP, Regular Expressions, 5
Text Normalisation, Edit Distance.

N-gram Language Models-N-Grams, Evaluating Language Models,


Generalisation and Zeros, Smoothing, Kneser-Ney Smoothing, The 5
Web and Stupid Backo , Perplexity’s Relation to Entropy.
INTERNAL TEST 1 (Module I)

II Neural Networks and Neural Language Models-Units, Feed-Forward


Neural Networks, Training Neural Nets, Neural Language Models.
4

Vector Semantics and Embeddings-Lexical Semantics, Vector


Semantics, Words and Vectors, Cosine for measuring similarity, TF-IDF:
Weighing terms in the vector, Applications of the tf-idf vector
model,Word2vec, Visualizing Embeddings ,Semantic properties of 7
embeddings, Bias and Embeddings, Evaluating Vector Models.
INTERNAL TEST 2 (Module II)

APJ Abdul Kalam Technological Universit


Page 46 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
fi
ff
y

III Sentiment Classi cation – What is sentiment classi cation. Machine


Learning for Sentiment Classi cation - Training the Classi er (Naive
Bayes, Logistic Regression, Support Vector Machine, Decision Tree, 6
Random Forest), Optimising for Sentiment Analysis - Other text
classi cation tasks – Evaluation of classi cation models: Precision,
Recall, F-measure, Test sets and Cross-validation, Statistical
Signi cance Testing.

Part-of-Speech Tagging-English Word Classes, The Penn Treebank


Part-of-Speech Tagset, Part-of-Speech Tagging, HMM Part-of-Speech
Tagging, Maximum Entropy Markov Models, Bi-directionality, Part-of-
Speech Tagging for Morphological Rich Languages. Information 6
Extraction-Named Entity Recognition, Relation Extraction, Extracting
Times, Extracting Events and their Times, Template Filling.
IV Sequence Processing with Recurrent Networks-Simple Recurrent
Neural Networks, Applications of Recurrent Neural Networks, Deep 5
Networks: Stacked and Bidirectional RNNs, Managing Context in
RNNs: LSTMs and GRUs, Words, Subwords and Characters

Neural Language Models and Generation Revisited, Encoder-Decoder


Networks, Attention, Applications of Encoder-Decoder Networks. Case
4
study: Machine translation, Question Answering

END SEMESTER EXAM (All Modules)


References:


1. Dan Jurafsky and James H. Martin. Speech and Language Processing (3rd ed)

2. Manning C, Schuetze H. Foundations of Statistical Natural Language Processing, MIT Press

3. James Allen, "Natural Language Understanding", 2/E, Addison-Wesley, 1994

4. Steven Bird, Natural Language Processing with Python, 1st Edition, O'Reilly, 2009

5. Jacob Perkins, Python Text Processing with NLTK 2.0 Cookbook, Packt Publishing, 2010


APJ Abdul Kalam Technological Universit


Page 47 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
fi
fi
y

fi
fi
fi
fi
Course Year of
Course Name L-T-P : C
Code Introduction

06AD6452 Machine Learning Models and Storage Management 3-0-0 : 3 2020

Course Objectives:


1. To familiarise the conventional and non-conventional methods of Machine Learning.

2. To familiarise and manage data organisation and access for di erent scales of processing.

3. Familiarise the lifecycle of machine learning process in normal and parallelised scenarios.
SYLLABUS:
Machine Learning through queries, Query optimisation, Execution strategies, Hardware
accelerators, Data Access, Resource management, ML life cycle, Parallel machines.
Course Outcomes:


Students should be able to

1. Apply database query languages to perform Machine Learning tasks.

2. Understand performance improvement of machine learning processes using hardware, data,


and resource acceleration.

3. Perform ML tasks by following standardised life cycle stages.

4. Understand how machine learning can be parallelised using di erent data structures.
Module Course Content Hours

I Machine Learning through database queries and UDFs, Sampling


based methods, Multi-table ML, Learning over joins, Statistical
relational learning, Deeper integration and specialised DBMSs. Scope 10
of optimization, planning rewrites, automatic operator fusion. Case
Study : Google Big Query
INTERNAL TEST 1 (Module I)

II Execution strategies - Data parallel and task parallel execution, 6


Model-parallel execution, Hybrid strategies, Hardware accelerators.
Case Study : GPU, TPU

Data Access - Caching and Bu er pool management, Data


compression, NUMA awareness, Indexing. 5

INTERNAL TEST 2 (Module II)

III Resource provisioning, scheduling, and con guration. Failure 5


management and transient resources. Case Study : AutoML

Managing ML life cycle, Data sourcing and cleaning, Feature


engineering and Deep learning, Model selection, management, and
deployment. Benchmarking ML systems. 5

IV Parallel machine models and pseudocode, Parallel algorithm


analysis, Processing in parallel - arrays, linked lists, queue-like
structures. Unbounded arrays. Hash tables and associative arrays, 9
Universal hashing, probing, perfect hashing, parallel hashing. Case
Study : IBM Parallel Machine Learning Toolbox.
END SEMESTER EXAM (All Modules)

APJ Abdul Kalam Technological Universit


Page 48 of 73 Cluster 6 | M. Tech Programme in AI & DS
y

ff
fi
ff
ff
References:

1. Matthias Boehm, Arun Kumar, Jun Yang, “Data Management in Machine Learning Systems”, Morgan
and Claypool

2. Peter Sanders, Kurt Mehlhorn, Martin Dietzfelbinger, Roman Dementiev, “Sequential and Parallel
Algorithms and Data Structures: The Basic Toolbox”, Springer 2019.

3. Ron Bekkerman, Mihkail Bilenko, John Langford, “Scaling up Machine Learning - Parallel and
Distributed Approaches”, Cambridge University Press.

APJ Abdul Kalam Technological Universit


Page 49 of 73 Cluster 6 | M. Tech Programme in AI & DS
y

Course Year of
Course Name L-T-P : C
Code Introduction

06AD6062 Mini Project 0-0-4 : 2 2020

Course Objectives:


1. To develop practical ability and knowledge about tools/techniques in order to solve the actual
problems related to the industry, academic institutions or similar area.

Course Outcomes:


Student should be able to

1. Identify and solve various problems associated with designing and implementing a intelligent
system or application. 

2. Test the designed system or application.
Syllabus 


Students can take up any application level/system level experimental design / implementation
tasks of relatively minor intensity and scope as compared to the major-project, pertaining to a
relevant domain of study. Projects can be chosen either from the list provided by the faculty or
in the eld of interest of the student. At the end of each phase, presentation and demonstration
of the project should be conducted, which will be evaluated by a panel of examiners. A detailed
project report duly approved by the guide in the prescribed format should be submitted by the
student for nal evaluation.

Publishing the work in Conference Proceedings/ Journals with National/ International status with
the consent of the guide will carry an additional weightage in the review process.

APJ Abdul Kalam Technological Universit


Page 50 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
fi
y

Course Year of
Course Name L-T-P : C
Code Introduction

06AD6072 Deep Learning Lab 0-0-3 : 1 2020

Course Objectives:


1. Implement the various deep learning algorithms in Python.

2. Learn to work with di erent deep learning frameworks like Keras, Tensor ow, PyTorch, Ca e
etc.
Course Outcomes:


Student should attain

1. Expert knowledge in solving real world problems using state of art deep learning techniques
References:


1. Francois Chollet, “Deep learning with Python” – Manning Publications.

Sl No List of Experiments

1 Basic image processing operations : Histogram equalization, thresholding, edge detection, data
augmentation, morphological operations

2 Implement SVM/Softmax classi er for CIFAR-10 dataset: (i) using KNN, (ii) using 3 layer neural
network

3 Study the e ect of batch normalisation and dropout in neural network classi er

4 Familiarisation of image labelling tools for object detection, segmentation

5 Image segmentation using Mask RCNN, UNet, SegNet

6 Object detection with single-stage and two-stage detectors (Yolo, SSD, FRCNN, etc.)

7 Image Captioning with Vanilla RNNs

8 Image Captioning with LSTMs

9 Network Visualisation: Saliency maps, Class Visualisation

10 Generative Adversarial Networks

11 Chatbot using bi-directional LSTMs

12 Familiarisation of cloud based computing like Google colab

Nine experiments to complete mandatory

APJ Abdul Kalam Technological Universit


Page 51 of 73 Cluster 6 | M. Tech Programme in AI & DS
ff
ff
y

fi
fi
fl
ff
APJ Abdul Kalam Technological University
Master of Technology – Course Plan

SEMESTER III

M. Tech Programme in
Arti cial Intelligence and Data Science

APJ Abdul Kalam Technological Universit


Page 52 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
y



Course Year of
Course Name L-T-P : C
Code Introduction

06AD7111 Arti cial Intelligence in Cyber Security 3-0-0 : 3 2020

Course Objectives:


1. To equip students realise the scope of arti cial intelligence in preventing security threats

2. To automate the process of detection using arti cial intelligence tools

3. To give an overview to the intrusion techniques

SYLLABUS:
Time Series analysis, Time series trends, Anomaly detection, Statistical and machine learning
approaches, Heuristics, Intrusion management.
Course Outcomes:


Student should be able to

1. Deploy arti cial intelligence based solutions for preventing cyber attacks

2. Understand the basic underlying architecture used for intrusion detection

3. Understand the heuristic methods used for cyber security
Module Course Content Hours
I Time series analysis, Stochastic time series model, ANN time series 5
model, Support Vector time series models, Time series decomposition,
Time series analysis in cybersecurity.

Time series trends and seasonal spikes, Predicting DDoS attacks -


ARMA, ARIMA, ARFIMA. Voting ensemble 5

INTERNAL TEST 1 (Module I)

II Using data science to catch email fraud and spam, Anomaly detection
using K-means, Using windows logs and active directory data.
Decision tree and Context-based malicious event detection.
4

Statistical and machine learning approaches to detection of


attacks on computers - Techniques for studying the Internet and
estimating the number and severity of attacks, network based attacks,
host based attacks. Statistical pattern recognition for detection and 6
classi cation of attacks, and techniques for visualizing network data,
etc.
INTERNAL TEST 2 (Module II)

III Using heuristics to detect malicious pages, Using machine learning,


logistic regression, and SVM to detect malicious URLs. Multiclass
5
classi cation to detect malicious URLs.

Levenshtein distance to di erentiate malicious URLs from others. Using


TensorFlow for intrusion detection. Machine learning to detect nancial
fraud - imbalanced data and credit card frauds, managing under-
sampled data for logistic regression. Adam gradient optimiser for deep 5
learning. Feature extraction and cosine similarity to quantify bad
passwords.

APJ Abdul Kalam Technological Universit


Page 53 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
fi
fi
fi
ff
y

fi
fi
fi
IV Overview of intrusions, system intrusion process, dangers of system
intrusions, history and state of the art of intrusion detection systems
(IDSs): anomaly detection, misuse detection, types of IDS: Network-
Based IDS. Host-Based IDS, Hybrid IDS, Intrusion Prevention Systems
10
(IPS): Network-Based IPS, Host-Based IPS, Intrusion Detection Tools,
the limitations and open problems of intrusion detection systems,
advanced persistent threats, case studies of intrusion detection
systems against real-world threats and malware.
END SEMESTER EXAM (All Modules)
References:


1. Soma Halder, Sinan Ozdemir, “Hands-on Machine Learning for Cybersecurity”, Packt
Publishing.

2. Roberto Di Pietro, Luigi V. Mancini, Intrusion Detection System, Springer ,2008

3. Anderson, Ross (2001). Security Engineering: A Guide to Building Dependable Distributed


Systems. New York: John Wiley & Sons. pp. 387–388. ISBN 978-0-471-38922-4.

4. Anderson, James P., "Computer Security Threat Monitoring and Surveillance," Washing, PA,
James P. Anderson Co., 1980.

APJ Abdul Kalam Technological Universit


Page 54 of 73 Cluster 6 | M. Tech Programme in AI & DS
y

Course Year of
Course Name L-T-P : C
Code Introduction

06AD7211 Game Theory in Arti cial Intelligence 3-0-0 : 3 2020

Course Objectives:

1. To introduce how human behaviour can be modelled using game theory principles for arti cial
intelligence

2. To learn various ways game theory helps in di erent learning mechanisms 



3. To introduce how game theory can be used to produce novel and accurate data for data
science problems
SYLLABUS:
Game theory, Nash Equilibrium, Cooperative games, Multi-agent AI, Imitation and reinforcement
learning, Adversary training.
Course Outcomes:


Student should be able to

1. Understand behavioural game theory for arti cial intelligence domain 

2. Learn the concepts of game theory for learning techniques in arti cial intelligence 

3. Apply game theoretic principles for dealing data for data science 

4. Model modern problems in AI and DS using game theory 

5. Implement game-theoretic solutions for AI and DS
Module Course Content Hours
I Introduction to Game Theory - Cooperative vs Non-Cooperative
Games, Symmetric vs Asymmetric Games, Perfect vs Imperfect 6
Information Games, Simultaneous vs Sequential Games, Zero-Sum vs
Non-Zero Sum Games, Nash Equilibrium, Inverse Game Theory

Two-person cooperative games without transferable payo s, N-


person cooperative games, Decisions under risk and uncertainty, 5
Decisions in con icts against p-intelligent players, Utility theory
INTERNAL TEST 1 (Module I)

II Multi-agent AI systems, Agent Architectures and Hierarchical control,


Multiagent framework, Representation of Games, Computing strategies
with perfect information, Planning under certainty, Partially observable
multi agent reasoning, Reasoning under uncertainty, Group decision 10
making, Mechanism design, Learning Belief networks, Ontologies and
Knowledge-based systems.

INTERNAL TEST 2 (Module II)

III Imitation and Reinforcement learning, Multi-agent Reinforcement


learning, Markov Decision process, Deep Q-learning, Imitation learning
with Dagger algorithm, Multi-arm bandits, Monte Carlo methods, 10
Temporal Di erence Learning, Policy Gradient methods. Case Study :
Personalised Web Services.

IV Adversary training, GANs, Generative Models - HMM, RBM,


Discriminative model - SVM, Attacks on Machine Learning, Conditional
GAN, DCGAN, InfoGAN, Stack GAN, Wasserstein GAN, Virtual batch 10
normalization.

APJ Abdul Kalam Technological Universit


Page 55 of 73 Cluster 6 | M. Tech Programme in AI & DS
ff
fl
fi
y

fi
ff
ff
fi
fi
END SEMESTER EXAM (All Modules)
References:

1. Maňas, M., “Games and Economic Decisions”, SNTL, Praha, 1998

2. Morris, P., “Introduction to Game Theory”, Springer Verlag, New York, 1994

3. David L. Poole, Alan K. Mackworth, “Arti cial Intelligence: Foundations of Computational


Agents”, Cambridge University Press

4. Andrea Lonza, “Reinforcement Learning Algorithms with Python”, Packt Publishing

5. Richard S. Sutton, Andrew G. Barto, “Reinforcement Learning: An Introduction”, MIT Press

6. Yevgeniy Vorobeychik, Murat Kantarcioglu, “Adversarial Machine Learning”, Morgan &


Claypool.

7. Navin K Manaswi, “Generative Adversarial Networks with Industrial Use Cases”, BPB.

APJ Abdul Kalam Technological Universit


Page 56 of 73 Cluster 6 | M. Tech Programme in AI & DS
y

fi
Course Year of
Course Name L-T-P : C
Code Introduction

06AD7311 Image and Video Analytics 3-0-0 : 3 2020

Course Objectives:


1. Familiarise the principles behind processing of video, audio, and image signals

2. Use di erent analytical models and loss functions for multimedia analytics
SYLLABUS:
Audio sampling, Image color spaces, Loss functions, Digital Image Segmentation, Object
classi cation, Object tracking, Speech and handwriting recognition.
Course Outcomes:


Student should be able to

1. Represent multimedia content using appropriate coding methods and quantisation.

2. Understand loss functions involved in quantised signals.

3. Apply segmentation, object detection, and ltering methods for real world applications.

4. Apply multimedia analytics in domains of speech recognition, handwriting recognition, and


object detection.
Module Course Content Hours
I Audio acquisition - Sampling and aliasing, Sampling theorem, Linear
quantization, Nonuniform scalar quantization, Time-domain audio 7
processing, Linear predictive coding. Image color spaces, image
representation, formats, and descriptors. Video principles and MPEG
standard.

Loss function, Zero-one loss function, Quantization Error Minimization, 4


Vector quantisation, Neural Gas and Topology Representing Network.
INTERNAL TEST 1 (Module I)

II Digital Image Segmentation - Classi cation of segmentation


techniques, Edge detection, Edge linking, Thresholding, Region 7
growing, Region splitting and merging, Watershed based
segmentation. Shadow detection and removal. Image processing using
OpenCV - blending, smoothing, and reshaping.

Object Classi cation - Shape based object classi cation, Motion


based object classi cation, Viola Jones Object Detection Framework, 4
Object classi cation using CNNs, use of RCNN for object classi cation.
INTERNAL TEST 2 (Module II)

III Video Object tracking, Temporal models, Kalman lter, Region based
tracking, Contour based tracking, Feature based tracking, Model based 9
tracking, Particle ltering, Models for shape, style, and identity.
IV Speech and Handwriting recognition - HMM, Lexicon selection, N-
gram model performance. Video segmentation - Shot boundary
detection, Keyframe extraction, Hand pose colour-based recognition.
10
Baggage exchange detection - Use of GMM, Tracking using Kalman
lter, Object labelling. Classi cation of building images in video
sequences - Edge based recognition, changing region detection.
END SEMESTER EXAM (All Modules)
APJ Abdul Kalam Technological Universit
Page 57 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
fi
ff
fi
fi
fi
fi
y

fi
fi
fi
fi
fi
fi
References:


1. Francesco Camastra, Alessandro Vinciarelli, “Machine Learning for Audio, Image and Video
Analysis: Theory and Applications”, Springer 2015.

2. Maheshkumar H Kolekar, “Intelligent Video Surveillance Systems: An Algorithmic Approach”,


CRC Press.

3. Vesna Zeljkovic, “Video Surveillance Techniques and Technologies”, IGI Global

4. Himanshu Singh, “Practical Machine Learning and Image Processing”, APress, 2019

5. Simon J. D. Prince, “Computer Vision: Models, Learning, and Inference”, Cambridge


University Press

APJ Abdul Kalam Technological Universit


Page 58 of 73 Cluster 6 | M. Tech Programme in AI & DS
y

Course Year of
Course Name L-T-P : C
Code Introduction

06AD7411 Cloud Data Management 3-0-0 : 3 2020

Course Objectives:


1. Familiarise the di erent types of cloud infrastructures.

2. Manage cloud infrastructure in terms of organisation, scale, and security.

3. Appraise di erent cloud o erings based on replication and availability.


SYLLABUS:
Cloud infrastructures, Data security in cloud, Data location and control, Storage management
locations, Data security during mobility.
Course Outcomes:


Student should be able to

1. Demonstrate the concepts and technologies of Cloud Computing

2. Understand the security aspects associated with Cloud Computing

3. Demonstrate the virtual server component of Cloud Computing

4. Understand Cloud storage and usage monitoring along with security mechanism
Module Course Content Hours
I Cloud infrastructures; public, private, hybrid. Service provider
interfaces; Saas, Paas, Iaas. VDC environments; concept, planning and
design, business continuity and disaster recovery principles. Managing
11
VDC and cloud environments and infrastructures. Scalability and Cloud
Services- Large Scale Data Processing- Databases and Data Stores-
Data Archival.
INTERNAL TEST 1 (Module I)

II Data Security - Storage strategy and governance; security and


regulations. Designing secure solutions; the considerations and
implementations involved. Securing storage in virtualized and cloud 10
environments. Monitoring and management; security auditing and
SIEM.

INTERNAL TEST 2 (Module II)

III Data Location and Control - Architecture of storage, analysis and


planning. Storage network design considerations; NAS and FC SANs, 6
hybrid storage networking technologies (iSCSI, FCIP, FCoE), design for
storage virtualization in cloud computing, host system design
considerations.

Global storage management locations, scalability, operational


e ciency. Global storage distribution; terabytes to petabytes and
greater. Policy based information management; metadata attitudes; le 6
systems or object storage.
IV Securing data for transport, Designing backup/recovery solutions to
guarantee data availability in a virtualized environment. Design a
replication solution, local remote and advanced. Investigate Replication 9
in NAS and SAN environments. Data archiving solutions; analyzing
compliance and archiving design considerations.

END SEMESTER EXAM (All Modules)


APJ Abdul Kalam Technological Universit
Page 59 of 73 Cluster 6 | M. Tech Programme in AI & DS
ffi
ff
ff
y

ff
fi
References:

1. Greg Schulz, “Cloud and Virtual Data Storage Networking”, Auerbach Publications, 2011.

2. Marty Poniatowski, “Foundations of Green IT” Prentice Hall; 1 edition, 2009.

3. EMC, “Information Storage and Management” Wiley; 2 edition, 2012.

4. Volker Herminghaus, Albrecht Scriba, “Storage Management in Data Centers” Springer


2009.

5. Klaus Schmidt, “High Availability and Disaster Recovery” Springer 2006.

APJ Abdul Kalam Technological Universit


Page 60 of 73 Cluster 6 | M. Tech Programme in AI & DS
y

Course Year of
Course Name L-T-P : C
Code Introduction

06AD7121 Data Visualisation Techniques 3-0-0 : 3 2020

Course Objectives:


1. Familiarise how data can be presented to various stakeholders.

2. Identify peculiarities in data with the help of visualisation.

3. Design dashboards for easy understanding of underlying data.


SYLLABUS:
Visualization, Plotting in R, Visual analytics, Validation, Presentation to stakeholders, ggplot
library, Dashboards.
Course Outcomes:


Student should be able to

1. Understand the necessity of visualisation in data management.

2. Apply visual analytics principles to appropriately preprocess data for visualisation.

3. Use R functions to generate plots for given data.

4. Perform validation of visualisations based on type and purpose of data.

5. Create dashboards and drill-down methods for data visualisation.


Module Course Content Hours
I Introduction to visualization - the visualization pipeline, The Value of 5
Visualization, Data - Why Do Data Semantics and Types Matter, Data
Types, Dataset Types, Attribute Types, Semantics

Plotting in R - plot() function, Displaying multivariate data, Using


graphics parameters, Matrix plots, Exporting graphs. 6

INTERNAL TEST 1 (Module I)


II Visual Analytics - Optimal visualization types, Binning values,
Calculated elds, Table calculations, Level of Detail calculations.
5

Validation - Four Levels of Design, Angles of Attack, Threats and


Validation Approaches, Validation Examples, De ning Marks and
Channels, Using Marks and Channels, Channel E ectiveness, Relative 6
vs. Absolute Judgments.

INTERNAL TEST 2 (Module II)


III Presenting results to stakeholders, ggplot library in R - layers,
geoms, stats, positioning, annotations, scales, axes and legends,
10
facetting, autoplot and fortify (atleast one example of each case to be
done).
IV Dashboard development - Dashboard design principles, Dashboard
interactivity, Connected “drill-down” dashboards. Visualization case 9
studies - Textual data, Temporal data.

END SEMESTER EXAM (All Modules)

APJ Abdul Kalam Technological Universit


Page 61 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
y

ff
fi
References:


1. Few, Stephen, “Show Me the Numbers: Designing Tables and Graphs to Enlighten.” 2nd
Edition. Analytics Press 2012

2. Tamara Munzner, Visualization Analysis and Design (VAD), CRC press, 2014

3. Complete ggplot reference manual at https://ggplot2.tidyverse.org/reference/

4. Nina Zumel, John Mount "Practical Data Science with R··. Manning Publications. 2014

APJ Abdul Kalam Technological Universit


Page 62 of 73 Cluster 6 | M. Tech Programme in AI & DS
y

Course Year of
Course Name L-T-P : C
Code Introduction

06AD7221 Social Network Analysis 3-0-0 : 3 2020

Course Objectives:


1. To familiarise the components of the social network.

2. To model and visualise the social network.

3. To mine the users in the social network.

4. To familiarise the evolution of the social network.

5. To know the applications in real time systems.

SYLLABUS:
Web architecture, Visualizing online social networks, Modelling social networks, Aggregation
and reasoning, Evolution in social networks, Case Study.
Course Outcomes:


Student should be able to

1. Work on the internal components of the social network

2. Model and visualise the social network

3. Mine the behaviour of the users in the social network

4. Predict the possible next outcome of the social network


Module Course Content Hours
I Introduction to Web - Limitations of current Web – Development of
Semantic Web – Emergence of the Social Web – Statistical Properties
of Social Networks -Network analysis - Development of Social Network 10
Analysis - Key concepts and measures in network analysis - Discussion
networks - Blogs and online communities - Web-based networks.
INTERNAL TEST 1 (Module I)

II Visualizing Online Social Networks - A Taxonomy of Visualizations -


Graph Representation - Centrality- Clustering - Node-Edge Diagrams -
Visualizing Social Networks with Matrix- Based Representations-
Node-Link Diagrams - Hybrid Representations - Modelling and 11
aggregating social network data – Random Walks and their
Applications –Use of Hadoop and Map Reduce - Ontological
representation of social individuals and relationships.
INTERNAL TEST 2 (Module II)

III Aggregating and reasoning with social network data, Advanced


Representations – Extracting evolution of Web Community from a
Series of Web Archive - Detecting Communities in Social Networks -
10
Evaluating Communities – Core Methods for Community Detection &
Mining - Applications of Community Mining Algorithms - Node
Classi cation in Social Networks.

APJ Abdul Kalam Technological Universit


Page 63 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
y

IV Evolution in Social Networks – Framework - Tracing Smoothly


Evolving Communities - Models and Algorithms for Social In uence
Analysis - In uence Related Statistics - Social Similarity and In uence -
In uence Maximization in Viral Marketing - Algorithms and Systems for
Expert Location in Social Networks - Expert Location without Graph 11
Constraints - with Score Propagation – Expert Team Formation - Link
Prediction in Social Networks - Feature based Link Prediction –
Bayesian Probabilistic Models - Probabilistic Relational Models. Case
Study of Google Page Rank and Facebook Graphs.
END SEMESTER EXAM (All Modules)
References:


1. Ajith Abraham, Aboul Ella Hassanien, Václav Snášel, Computational Social Network
Analysis: Trends, Tools and Research Advances, Springer, 2012

2. Borko Furht, Handbook of Social Network Technologies and Applications, Springer, 1st
edition, 2011

3. Charu C. Aggarwal, Social Network Data Analytics‖, Springer; 2014

4. Giles, Mark Smith, John Yen, Advances in Social Network Mining and Analysis, Springer,
2010.

5. Guandong Xu , Yanchun Zhang and Lin Li, Web Mining and Social Networking – Techniques
and applications, Springer, 1st edition, 2012

6. Peter Mika, Social Networks and the Semantic Web, Springer, 1st edition, 2007.

7. Przemyslaw Kazienko, Nitesh Chawla, Applications of Social Media and Social Network
Analysis, Springer, 2015

APJ Abdul Kalam Technological Universit


Page 64 of 73 Cluster 6 | M. Tech Programme in AI & DS
fl
fl
y

fl
fl
Course Year of
Course Name L-T-P : C
Code Introduction

06AD7321 Text Mining 3-0-0 : 3 2020

Course Objectives:


1. Apply text processing techniques to prepare documents for statistical modelling 

2. Apply relevant machine learning models for analysing textual data and correctly interpreting
the results 

3. Use machine learning models for text prediction 

4. Evaluate the performance of machine learning models for textual data
SYLLABUS:
Basic NLP processes, Document representation, Text categorization and clustering, Topic
modeling, Document summarization, Sentiment analysis, Text visualization.
Course Outcomes:


Student should be able to

1. Describe basic concepts and methods in text mining, for example text representation, text
classi cation and clustering, and topic modelling
2. Use the text mining concepts and methods to model real-world problems into text mining
tasks, develop technical solutions, and evaluate the e ectiveness of the solutions.
3. Communicate text mining process, result, and major ndings to various audience including
both experts and laypersons.

Module Course Content Hours


I Basic techniques in natural language processing - tokenziation,
part-of-speech tagging, chunking, syntax parsing, named entity 6
recognition. Case study : Public NLP toolkits.

Document representation - representing unstructured text documents


with appropriate format and structure, automated text mining 5
algorithms.
INTERNAL TEST 1 (Module I)

II Text categorization - supervised text categorization algorithms, Naive 6


Bayes, kNN, Logistic Regression, SVM, Decision Trees.

Text clustering - connectivity-based (or hierarchical) clustering,


centroid-based (k-means) clustering. 5

INTERNAL TEST 2 (Module II)

III Topic modeling - general idea of topic modeling, basic topic models,
Probabilistic Latent Semantic Indexing, Latent Dirichlet Allocation 7
(LDA). Applications - classi cation, imagine annotation, collaborative
ltering, and hierarchical topical structure modeling.


Document summarization - Extraction- based summarization 3


methods.

APJ Abdul Kalam Technological Universit


Page 65 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
fi
y

fi

ff
fi

IV Sentiment analysis - concept, sentiment polarity prediction, review 6
mining, aspect identi cation.


Text visualization - introduction to mathematical and programming


tools. 3

END SEMESTER EXAM (All Modules)


References:


1. Charu C. Aggarwal and Cheng Xiang Zhai, “Mining Text Data”, Springer, 2012.

2. Daniel Jurafsky and James H Martin, “Speech & Language Processing”, Pearson Education
India, 2000.

3. Christopher D. Manning, Prabhakar Raghavan, and Hinrich Schuetze, “Introduction to


Information Retrieval”. Cambridge University Press, 2007.

APJ Abdul Kalam Technological Universit


Page 66 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
y

Course Year of
Course Name L-T-P : C
Code Introduction

06AD7421 Data Warehouse and Data Lakes 3-0-0 : 3 2020

Course Objectives:


1. Familiarise the need of data warehouse and data lake in the context of data mining.

2. Familiarise the use of OLAP and Data cubes for data mining.

3. Identify data management options for data lakes.

4. Integrate data lake with existing data APIs.


SYLLABUS:
Data mining, Data Warehouse, OLAP, Data cubes, Data lakes, Visual analytics, Integrations with
existing analytical frameworks.
Course Outcomes:


Student should be able to

1. Understand the process of mining insights from large data.

2. Apply data processing and structuring to organise data in warehouses.

3. Create dashboards for illustrating data and insights from data lakes.

4. Design APIs to integrate data lakes with existing data delivery methods.
Module Course Content Hours

I Introduction: Fundamentals of data mining, Data Mining


Functionalities, Classi cation of Data Mining systems, Data Mining
Task Primitives, Integration of a Data Mining System with a Database or
a Data Warehouse System, Major issues in Data Mining. Data 10
Preprocessing: Need for Preprocessing the Data, Data Cleaning, Data
Integration and Transformation, Data Reduction, Discretization and
Concept Hierarchy Generation.
INTERNAL TEST 1 (Module I)

II Data Warehouse and OLAP Technology for Data Mining: Data


Warehouse, Multidimensional Data Model, Data Warehouse
Architecture, Data Warehouse Implementation, Further Development of
Data Cube Technology, From Data Warehousing to Data Mining Data 11
Cube Computation and Data Generalization: E cient Methods for Data
Cube Computation, Further Development of Data Cube and OLAP
Technology, Attribute-Oriented Induction.
INTERNAL TEST 2 (Module II)

III Data Lake, Data lake maturity, Data puddles, Data ponds, Data lake
organisation - landing zone, production zone, work zone, sensitive
zone. Di erences between Data lake and Data Warehouse. Data lake
Business Reporting, Visual Analytics: De nition, concepts, Di erent 11
types of charts and graphs, Emergence of data visualization and visual
analytics. Lambda architecture driven data lake, Applied lambda for
data lake.
IV Integrations - Acquisition of batch data using Apache Sqoop,
Acquisition of stream data using Apache Flume, Messaging layer using
10
Apache Kafka, Data Store using Apache Hadoop, Indexed datastore
using ElasticSearch.
END SEMESTER EXAM (All Modules)

APJ Abdul Kalam Technological Universit


Page 67 of 73 Cluster 6 | M. Tech Programme in AI & DS
ff
fi
y

fi
ffi
ff
References:

1. Sam Aanhory, Dennis Murray , "Data Warehousing in the Real World”, Pearson Edn Asia.

2. Paulraj Ponnaiah, "Data Warehousing Fundamentals”, Wiley student Edition

3. Alex Gorelik, “The Enterprise Big Data Lake”, O’Reilly

4. Tomcy John, Pankaj Misra, “Data Lake for Enterprises”, Packt Publishing

5. Jiawei Han & Micheline Kamber, “Data Mining – Concepts and Techniques”, Morgan
Kaufmann Publishers, Elsevier, 2nd Edition, 2006.

6. U. Dinesh Kumar, “Business Analytics – The Science of Data Driven Decision Making”, Wiley
2017.

7. Ramesh Sharda, DursunDelen, Efraim Turban, “Business Intelligence: A Managerial


Perspective on Analytics”, Pearson, 3e.

APJ Abdul Kalam Technological Universit


Page 68 of 73 Cluster 6 | M. Tech Programme in AI & DS
y

Course Year of
Course Name L-T-P : C
Code Introduction

06AD7031 Seminar II 0-0-2 : 2 2020

Course Objectives:


1. To introduce the students to research, make them understand research papers and prepare
presentation material

2. To understand cutting edge technology in the chosen area



3. To improve oral communication skills through presentation

4. To prepare original technical write up on the presentation

Course Outcomes:


After completion of course, students will be able to:

1. Develop skills in doing literature survey, technical presentation and report preparation
2. Improve the pro ciency in English

3. Improve presentation skills

4. Improve analytical and reasoning ability

5. Improve technical writing skills

Syllabus:


The aim of this course is to introduce the student to research, and to acquaint him with the
process of presenting his work through seminars and technical reports. Students have to
register for the seminar and select a topic in consultation with any faculty member o ering
courses for the programme. The student is expected to do an extensive literature survey and
analysis in an area related to the area of specialisation. The study should preferably result in
design ideas, designs, algorithms, and theoretical contributions in the form of theorems and
proofs, new methods of proof, new techniques or heuristics with analytical studies,
implementations and analysis of results.

The presentation shall be of 30 minutes duration and a committee with the Head of the
Department as the chairman and two faculty members from the department as members
shall evaluate the seminar based on the coverage of the topic, presentation and ability to
answer the questions put forward by the committee.

Students shall individually prepare and submit a seminar report based on experimental
study / industrial training on the corresponding topic, in the prescribed format given by the
Department. The reference shall include standard journals (ACM/IEEE), conference
proceedings and equivalent documents, reputed magazines and textbooks, technical reports
and web based material, approved by the supervisor. The references shall be incorporated in
the report following IEEE standards re ecting the state-of-the-art in the topic selected.

APJ Abdul Kalam Technological Universit


Page 69 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
y

fl

ff
Course Year of
Course Name L-T-P : C
Code Introduction

06AD7041 Project Phase I 0-0-8 : 6 2020

Course Objectives:


1.To undertake research in an area related to the program of study.

2. To acquaint students to literature survey and design of a project.

Course Outcomes:


Student should be able to

1. Identify the topic, objectives and methodology to carry out the project. 

2. Finalise the project plan for their course project.
Syllabus:


Every student should carry out project, related to areas of Data Sciences, under the
supervision of a Supervisor(s). The project work shall commence in the third semester and
shall be completed by the end of fourth semester. Candidates are required to undertake a
suitable research project work; the topic shall be approved by a committee constituted by the
Head of the concerned Department. Every student will be required to present the topic at the
beginning of the Phase-I to illustrate the scope of the work and to nalize the topic. The third
semester includes the design phase and the fourth semester includes the implementation
and nal thesis submission.

The student should report the status of their progress weekly to the concerned supervisor.
Students should submit the project report at the end of the respective semesters, on dates
announced by the college/department. Project evaluation will be based on presentations,
viva voce, demonstration, review reports, design reports and nal thesis. Progress of the
project work is to be evaluated at the end of the third semester. For this a committee headed
by the head of the department with two other faculty members in the area of the project, of
which one shall be the project supervisor. If the project is done outside the college, the
external supervisor associated with the student will also be a member of the committee.

Normally students are expected to do the project within the college. However they are
permitted to do the project in an industry or in a government research institute under a
quali ed supervisor from that organization. This is only possible in the fourth semester and
the topic of investigation should be in line with the project part planned in the 3rd semester.
Student should apply for this through the project supervisor indicating the reason for this well
in advance, preferably at the beginning of the 3rd semester.

Project evaluation marks shall be as follows:-

Total marks for the Project: 150

In the 3rd Semester: Marks:50

Project Progress evaluation:

Progress evaluation by the Project Supervisor : 20 Marks

Presentation and evaluation by the committee : 30 Marks

APJ Abdul Kalam Technological Universit


Page 70 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
fi
y

fi
fi
APJ Abdul Kalam Technological University
Master of Technology – Course Plan

SEMESTER IV

M. Tech Programme in
Arti cial Intelligence and Data Science

APJ Abdul Kalam Technological Universit


Page 71 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
y



Course Year of
Course Name L-T-P : C
Code Introduction

06AD7012 Project Phase II 0-0-21 : 12 2020

Course Objectives:
1. To undertake research in an area related to the program of study.

2. To enable students to implement and deploy a system and carry out performance analysis.
Course Outcomes:

Student should be able to

1. Get a good exposure to a domain of interest. 



2. Get a good domain and experience to pursue future research activities.
Syllabus:
The Phase II work shall be based on the work in Phase I. Normally students are expected to
do the project within the college. However they are permitted to do the project in an industry
or in a government research institute under a quali ed supervisor from that organization; the
topic of investigation should be in line with the project part planned in the 3rd semester.
Student should apply for this through the project supervisor indicating the reason for this well
in advance, preferably at the beginning of the 3rd semester. This application is to be vetted
by a departmental committee constituted for the same by the Principal and based on the
recommendation of the committee the student is permitted to do the project outside the
college. The same committee should ensure the progress of the work periodically and keep a
record of this. The application for this shall include the following:-

Topic of the Project, Project work plan in the 3rdSemester, Reason for doing the project
outside, Institution/Organization where the project is to be done, External Supervisor Name,
Designation, Quali cation and Experience, Letter of consent of the External Supervisor as
well as from the organization.

Final evaluation of the project will be taken up only on completion of the project in the fourth
semester. This shall be done by a committee constituted for the purpose by the principal of
the college. The concerned head of the department shall be the chairman of this committee.
It shall have two senior faculty members from the same department, project supervisor and
the external supervisor, if any, of the student and an external expert either from an academic/
R&D organization or from Industry as members. Final project grading shall take into account
the progress evaluation done in the third semester and the project evaluation in the fourth
semester. If the quantum of work done by the candidate is found to be unsatisfactory, the
committee may extend the duration of the project up to one more semester, giving reasons
for this in writing to the student. Normally further extension will not be granted and there shall
be no provision to register again for the project.

Project work is to be evaluated both in the third and the fourth semesters. Based on these
evaluations the grade is nalized in the fourth semester.

Project evaluation marks shall be as follows:-

Total marks for the Project: 150

In the 4th Semester: Marks:100

Project evaluation by the supervisor/s : 30 Marks

Presentation& evaluation by the Committee : 40 Marks

Evaluation by the External expert : 30 Marks

Students are required to publish their work in reputed national/ International Journals/
Conference Proceedings etc which will carry weightage in nal marks.

APJ Abdul Kalam Technological Universit


Page 72 of 73 Cluster 6 | M. Tech Programme in AI & DS
fi
fi
y

fi
fi
APJ Abdul Kalam Technological Universit
Page 73 of 73 Cluster 6 | M. Tech Programme in AI & DS
y

You might also like