| Course Code: CS5301 Course Name: Machine Learning Prerequisites: CS561 Syllabus: Definitions, goals and history of Machine Learning. Taxonomies of methods and research paradigms. Knowledge-level vs. symbol-level learning. Major approaches of learning: Inductive concept acquisition (version-space, ID3, and AQ algorithms); inductive bias, minimum description length principle. Formal models of learnability. learning in the limit. PAC learnability. Ockham's razor. Learning by observation and discovery (e.g., conceptual clustering in CLUSTER and COBWEB) Scientific and mathematical discovery (e.g., AM and BACON) Explanation-based learning: macro-operators (STRIPS), explanation generalization (as in EBG, EGGS, and SOAR); Connectionist (i.e. neural network) learning (perception and back-propagation), Analogy and case-based reasoning (exemplars, structure mapping). Texts: 1. J. Shavlik and T. Dietterich (Ed), Readings in Machine Learning, Morgan Kaufmann, 1990. 2. P. Langley, Elements of Machine Learning, Morgan Kaufmann, 1995. |
| Course Code: CS5301 Course Name: Machine Learning Prerequisites: NIL Syllabus: Introduction to machine learning: Learning paradigms and applications, supervised vs unsupervised learning, training and test data, generalization and overfitting, model evaluation metrics; Models for Regression: Linear regression, least squares estimation, basis function expansion, bias-variance trade-off, regularization methods including ridge regression; Models for Classification: Classification framework, decision boundaries, logistic regression, multiclass softmax regression, discriminative vs generative classification, perceptron learning algorithm, regularization and model selection, KNN classification, Decision trees; Neural Networks: Multilayer perceptrons, network architectures, activation functions, forward propagation, backpropagation learning algorithm, gradient descent optimization, regularization and generalization in neural networks; Kernel Methods: Kernel functions, kernel trick, feature space transformations, support vector machines, margin maximization, kernel regression, kernel classification, sparse kernel machines and relevance vector machines; Unsupervised Learning and Dimensionality Reduction: Clustering methods including k-means, vector quantization, principal component analysis (PCA), probabilistic PCA; Introduction to Reinforcement Learning; Texts: 1. Christopher M. Bishop, Pattern Recognition and Machine Learning, Springer NP Exclusive (CBS), 2009. 2. Kevin P. Murphy, Machine Learning: A Probabilistic Perspective, The MIT Press, 2012. 3. Tom M. Mitchell, Machine Learning, McGraw-Hill Education, 2017 |