DS 303 - Introduction to Machine Learning

Course content
  • Introduction to machine learning: What is learning, learning objectives, data needed.
  • Supervised Learning.
  • Bayesian inference and learning: Inference, naïve Bayes.
  • Measures of success and loss functions.
  • Generalization and model complexity, bias-variance tradeoff. Training, validation, and testing.
  • Introduction to convex optimization. Convergence and training time. Objective functions for classification, regression, and ranking.
  • Linear regression, Perceptron and logistic regression.
  • MLP and backpropagation.
  • Deep learning, CNN and RNN.
  • SVM, support vector regression, increase in dimensionality through simple kernels. Decision trees.
  • Role of randomization and model combination, bagging and boosting.
  • Unsupervised Learning. Clustering criteria, K-means, DB-scan.
  • Kernel Density estimation. EM-algorithm for mixture of Gaussians.
  • Dimensionality reduction using PCA and Kernel-PCA.
  • Other Topics: Overview of Reinforcement Learning. Bias and Ethics in ML.
  • Optional topics: Introduction to one or more of the following topics: Active/Transfer Learning, Bootstrapping, Semi-supervised learning, Generative and probabilistic graphical models, Online/incremental learning.
References
  • Pattern Recognition and MachineLearning, by Christopher Bishop,Springer 2011
  • The Elements of Statistical Learning:Data Mining, Inference, and Prediction,Second Edition, by Trevor Hastie andRobert Tibshirani (Springer Series inStatistics) 2016
  • Supplementary material available online,e.g. Dive into Deep Learning by AstonZhang, Zack C. Lipton, Mu Li andAlexander Smola, 2020 (https://d2l.ai)
Pre-requisite : N/A
Total credits : 6 credits - Lecture
Type : Core Course
Duration : Full Semester
Name(s) of other Academic units to whom the course may be relevant : N/A