Mumbai University · Final Year Information Technology · Rev-2019 'C' Scheme
AI and DS II
ITC701 · Semester VII · 3 credits · Internal tests 20 (average of two) + end-semester 80
End-semester paper (3 hours): Six questions of 20 marks; Q1 is compulsory and you answer four. Each module's share of the paper follows its lecture hours.
Each module's share of the paper follows its lecture hours, the only IT final-year course that says so.
Module 1: Uncertainty
4 hoursUncertainty in AI; inference using full joint distributions; Bayes' theorem; semantics of and inference in Bayesian networks; decision theory; Markov decision processes.
Examined in the 80-mark end-semester paper (about 11% of it, by lecture hours); the class tests cover what has been taught by then
- Uncertainty in AI
- Inference using full joint distributions
- Bayes' theorem in AI
- Bayesian networks and inference
- Decision theory and Markov decision processes
Module 2: Cognitive Computing
6 hoursFoundations and design principles of cognitive systems; NLP in cognitive systems; taxonomies and ontologies; advanced analytics; building a cognitive application.
Examined in the 80-mark end-semester paper (about 16% of it, by lecture hours); the class tests cover what has been taught by then
- Foundations of cognitive computing
- Design principles for cognitive systems
- NLP in cognitive systems
- Taxonomies and ontologies
- Building a cognitive application
Module 3: Fuzzy Logic and its Applications
7 hoursFuzzy sets, properties and operations; membership functions; fuzzy relations; max–min and max–product composition; defuzzification; Mamdani fuzzy controllers for showers, washing machines and water purifiers.
Examined in the 80-mark end-semester paper (about 19% of it, by lecture hours); the class tests cover what has been taught by then
- Fuzzy sets and their operations
- Membership functions
- Fuzzy relations and composition
- Defuzzification methods
- Mamdani fuzzy controllers
Module 4: Introduction to Deep Learning
8 hoursANNs; ML versus deep learning; CNN components, properties and architectures; simple and deep RNNs and LSTM; autoencoders; applications.
Examined in the 80-mark end-semester paper (about 22% of it, by lecture hours); the class tests cover what has been taught by then
- Deep learning versus machine learning
- Convolutional neural networks
- Recurrent neural networks and LSTM
- Autoencoders
Module 5: Advanced ML Classification
6 hoursEnsembles: bagging, boosting and random forests; class-imbalanced data; classifier metrics; holdout, random subsampling, cross-validation and bootstrap; model selection by significance tests; cost–benefit and ROC comparison.
Examined in the 80-mark end-semester paper (about 16% of it, by lecture hours); the class tests cover what has been taught by then
- Ensemble methods: bagging, boosting and random forests
- Class-imbalanced data
- Evaluating classifiers: holdout, cross-validation and bootstrap
- Comparing classifiers with ROC curves and significance tests
Module 6: Trends in Data Science
6 hoursData science applications and case studies; data science for text, image, video and audio; multimodal applications.
Examined in the 80-mark end-semester paper (about 16% of it, by lecture hours); the class tests cover what has been taught by then
- Data science applications and case studies
- Data science for text, images, video and audio
- Multimodal data science
Module contents are summarised from the official Mumbai University syllabus . AiCanCode.org is not affiliated with the University of Mumbai.