Home / Semester 5 / CSE472
Back to Subjects

CSE472 — Deep Learning For Natural Language Processing

Practice

Course Overview

CSE472 3 Credits L:2 T:0 P:2 Minor Machine Learning
Download Syllabus

This subject covers the foundations and linguistic aspects of natural language processing, text preprocessing, word embeddings, deep learning sequence models, sequence-to-sequence architectures, attention mechanisms, transformers, and pretrained language models. It also addresses generative NLP and large language models, including text generation, evaluation, explainability, and hallucination.

Study Units

Unit 1

Foundations of NLP and Text Processing

Notes MCQ Subjective
Unit 2

Word Embeddings and Vector Representations

Notes MCQ Subjective
Unit 3

Deep Learning Sequence Models for NLP

Notes MCQ Subjective
Unit 4

Sequence-to-Sequence Models and Attention Mechanisms

Notes MCQ Subjective
Unit 5

Transformers and Pretrained Language Models

Notes MCQ Subjective
Unit 6

Generative NLP and LLMs

Notes MCQ Subjective

Continuous Assessment

3 components

The first CA is mandatory; the best 1 of the remaining 2 also counts.
Project 66.66%

Students are required to carry out a comprehensive NLP project covering the entire course syllabus, involving problem identification, dataset selection, text preprocessing, model design and implementation using deep learning and transformer-based approaches, performance evaluation using suitable metrics, analysis of results, documentation, and presentation.

Week 3 / 12

Rubric
Technical Execution & Skill Demonstration: 10 marks; Innovation & Project Report: 10 marks; Project Presentation & Viva (Q&A): 10 marks.

Test - Code based 33.34%

An objective-type individual assessment covering Units I-II of Deep Learning for NLP. It includes MCQs, numerical problems, and basic code-based questions on NLP foundations, word embeddings, and deep learning models, and evaluates conceptual clarity, analytical skills, and application of core NLP techniques.

Week 4 / 5

Rubric
The objective of this assessment is to evaluate students' understanding of fundamental concepts in Natural Language Processing covered in Units I-II.

Test 33.34%

Subjective test covering syllabus from unit1 to unit4.

Week 11 / 12

Rubric
To test the subject knowledge of the students and provide opportunity to improve their CA performance in case of low marks or any missed CA.

Exams & Practice

End Term Examination

Final semester comprehensive evaluation

50%

Type: Examination

All MCQ

CSE472 - FAQs

How many units are in CSE472?

CSE472 has 6 units. Each unit includes detailed notes and MCQ practice questions.

What exam resources are available for CSE472?

Unit-wise notes and MCQ practice are available. Exam resources coming soon.

How to prepare for CSE472 exams?

Study each unit's notes thoroughly, practice MCQs to test understanding, and attempt mock tests before exams. Focus on important topics and previous year questions.