AI & Machine LearningSemester 3-2 syllabus

The second half of third year deepens the AI&ML core with Natural Language Processing, Deep Learning, and Data Visualization, alongside two further Professional Elective slots and a second Open Elective. Dedicated labs give hands-on practice in deep learning frameworks and visualization tooling, while Soft Skills and a Technical Paper Writing & IPR audit course build communication and research-documentation ability. A mandatory industry internship or mini-project follows over the summer.

Subjects

Natural Language Processing

• Unit 1: Origins and challenges of NLP, including grammar-based and statistical language modeling.
• Unit 2: Word-level analysis — N-gram models, their evaluation, and smoothing techniques.
• Unit 3: Syntactic analysis using context-free grammars and English grammar rules.
• Unit 4: Semantics and pragmatics — representation requirements and first-order logic approaches.
• Unit 5: Discourse analysis and lexical resource construction.

L-T-P: 3-0-0, 3 credits

Deep Learning

• Unit 1: Biological inspiration for neural computation and the McCulloch-Pitts unit.
• Unit 2: Feedforward networks, gradient descent, and backpropagation.
• Unit 3: Modern optimization methods for more effective neural network training.
• Unit 4: Recurrent neural networks, backpropagation through time, and LSTM units.
• Unit 5: Recent developments — variational autoencoders, transformers, and GPT-style applications across vision and language tasks.

L-T-P: 3-0-0, 3 credits

Data Visualization

• Unit 1: What visualization is, its history, and its relationship to related disciplines.
• Unit 2: Creating visual representations — reference models and visual mapping.
• Unit 3: Classifying visualization systems and interaction techniques, including common pitfalls like misleading charts.
• Unit 4: Visualizing groups, trees, graphs, clusters, and networks.
• Unit 5: Visualizing volumetric data, vector fields, and simulation processes.

L-T-P: 3-0-0, 3 credits

Professional Elective-II options:

• Software Testing Methodologies — covers the purpose of testing, transaction-flow testing, path-based testing techniques, state-graph-based transition testing, and graph-matrix approaches to test design.
• Cryptography & Network Security — covers core security principles, symmetric cryptography’s algebraic foundations, block ciphers (DES, AES, Blowfish), cryptographic hash functions, and transport/web-level security.
• DevOps — covers the SDLC-to-DevOps transition, source-code management with Git, build automation and continuous integration, continuous delivery practices, and configuration management with Ansible.
• Recommender Systems — covers recommender-system fundamentals, collaborative filtering, content-based recommendation, hybrid approaches, and methods for evaluating recommender quality.

L-T-P: 3-0-0, 3 credits (one option selected)

Professional Elective-III options:

• Software Project Management — covers conventional (waterfall) project management, lifecycle phases, model-based software architectures, project organization structures, and Agile/Scrum adoption.
• Mobile Adhoc Networks — covers ad hoc and cellular network fundamentals, MANET routing protocols, ad hoc network security, wireless sensor network basics, and WSN security/key management.
• Computer Vision — covers camera models and radiometry, linear filters and convolution, multi-view geometry and stereopsis, model-fitting segmentation (Hough transform), and geometric camera calibration.
• NoSQL Databases — covers the four major NoSQL database types, comparisons with relational databases, key/value and document stores (MongoDB), column-oriented stores (HBase), and Riak-based key/value databases.

L-T-P: 3-0-0, 3 credits (one option selected)

Open Elective-II is drawn from the cross-department elective pool in the same way as Open Elective-I.

Deep Learning Lab

• Building multi-layer perceptrons for MNIST digit classification.
• Designing networks for binary and multi-class text classification (IMDB, Reuters datasets).
• Predicting housing prices with a regression network on the Boston Housing dataset.
• Building convolutional neural networks for digit and image classification, including transfer learning with VGG16.
• Implementing word embeddings and a recurrent neural network for movie-review sentiment classification.

L-T-P: 0-0-3, 1.5 credits

Data Visualization Lab

• Visualizing datasets with histograms and line charts.
• Building bar charts and box plots across multiple datasets.
• Creating scatter plots, mosaic plots, and multi-variable scatter matrices.
• Producing map-based visualizations and heatmaps.
• Building correlograms and 3D visualizations for multivariate data.

L-T-P: 0-0-3, 1.5 credits

Soft Skills

(Skill Enhancement Course)

• Analytical thinking and listening skills, including self-introduction and structured talks.
• Self-management skills — anger, stress, and time management.
• Standard operating methods for communication — grammar, tenses, and pronunciation.
• Job-oriented skills — group discussions and resume preparation.
• Interpersonal relationships — their importance, types, and influencing factors.

L-T-P: 0-1-2, 2 credits

Technical Paper Writing & IPR

(Audit Course)

• Introduction to technical report writing and sentence construction.
• Drafting reports and handling illustrations/graphics.
• Proofreading and summarization practice.
• Using word-processing tools for structured report elements like tables of contents.
• The nature of intellectual property — patents, designs, trademarks, and copyright.

L-T-P: 2-0-0 (non-credit audit course)

A mandatory industry internship or mini-project of 8 weeks’ duration is undertaken during the following summer vacation.

Semester total: 20-1-8 contact hours, 23 credits, with optional Minor and Honors-pool courses available as in the prior semester.