Machine Learning and Computational Intelligence
Syllabus
Lectures: 4 Teaching Hours per week
Tutorial: 0 Teaching Hours per week
Practical: 0 Teaching Hours per two weeks
Year: I
Part: I
Course Type: Core
Course Objectives
The objective of this course is to learn the fundamentals of pattern recognition and its relevance to classical and modern problems. It also talks about to identify where, when and how pattern recognition can be applied. The students will learn how to use pattern recognition in the real world. Students are introduced to recent applications of pattern recognition in the fields of science, Engineering, medicine, cognitive science and bioinformatics.
Course Outline
-
Introduction (5 hours) Artificial Intelligence, Machine learning, Computational Intelligence, History, Computational Intelligence Paradigms; Data and tools, review of statistics, training, validation and test data, theory of learning – feasibility of learning – error and noise – training versus testing, generalization bound – approximation-generalization tradeoff – bias and variance – learning curve
-
Feature Extraction (8 hours) Feature Selection: Data Preprocessing, Class Separability Measures, Feature Subset Selection, Bayesian Information Criterion Dimensionality Reduction: Basis Vectors, Singular Value Decomposition, Independent Component Analysis, Kernel PCA, Wavelets Additional Features and Template Matching: Texture, Shape and Size Characterization, Fractals, Features for Audio, Template Matching Using Dynamic Time Warping and Edit Distance
-
Bayes Decision Theory (5 Hrs) Discriminant Functions and Services, the Normal Distribution, Bayesian Classification, Estimating Probability Density Functions, Nearest Neighbor Rules, Bayesian Networks
-
Neural Networks (5 Hrs) Basic Neuron, Perceptrons, Hebbian Learning, Adaline Network , Application of Neural Networks
-
Classification (10Hrs) Linear Classifiers: the Perceptron Algorithm, Least-Squares Methods; Nonlinear Classifiers: Multilayer Perceptron’s, Back Propagation Algorithm, Decision Trees, Combinations of Classifiers, Boosting Context Dependent Classification; SVM and Other methods
-
Clustering (10Hrs) Sequential Algorithms, Hierarchical Algorithms, Functional Optimization-Based Clustering, Graph Clustering, Learning Clustering, Clustering High Dimensional Data, Subspace Clustering, Cluster Validity Measures
-
Deep Learning (9 Hrs) Introduction, Architectures, Hyperparameter tuning, Fundamentals on various networks like CNN, RNN, GRU,SNN, GAN e.t.c
-
Evolutionary Computation & Intelligence (6Hrs) Genetic Algorithm, Particle Swarm Optimization, Ant Colony Optimization, Fuzzy Logic
-
Performance Evaluation and Measures (4 Hrs) Classification accuracy, Confusion matrix Misclassification costs, Sensitivity and specificity, ROC curve, Recall and precision, box plot, confidence interval. Estimating Performance, Reliability of quality estimations Confidence interval, Cross validation Data visualization
Evaluation Scheme
a. Internal Examination
| Type | Weightage |
|---|---|
| Minor tests | 50% |
| Assignments | 50% |
b. Final Examination There will be five units of questions carrying 12 marks each. The question will cover all chapters of the syllabus. The evaluation Scheme will be as indicated in the table.
| S.N. | Chapter | Hours | Marks Distrubution |
|---|---|---|---|
| 1 | 1,8 | 5+6 | 12 |
| 2 | 2,9 | 8+4 | 12 |
| 3 | 3,5 | 5+10 | 12 |
| 4 | 4,7 | 5+9 | 12 |
| 5 | 6 | 10 | 12 |
| Total | 60 |
References
1. Engelbrecht AP. Computational intelligence: an introduction. John Wiley & Sons; 2007.
2. Pattern Recognition and Machine Learning, Christopher M. Bishop, Springer.
3. Introduction to Machine Learning with Python: A Guide for Data Scientists, 1st ed., O'Reilly Media
4. David Barber, Bayesian Reasoning and Machine learning, Oxford University press
5. Tom M Mitchell, Machine Learning, First edition, McGraw Hill Education
6. Simon Haykin, “Neural Networks: A Comprehensive Foundation”, PearsonEducation Asia, 2001.