Course Description: | Machine learning is defined as a set of methods that can automatically detect patterns in data, and then use the uncovered patterns to predict future data, or to perform other kinds of decision making under uncertainty. This course introduces the most important methods in statistical machine learning, especially the monitoring of learning methods, including perceptual machines, k-nearest neighbor method, naive Bayesian method, decision tree, logistic regression and maximum entropy model, support vector machine, EM algorithm, hidden Markov model and conditional random field. The course mainly includes: the introduction of three elements of statistical study, supervised learning, model evaluation and model selection, regularization and cross validation, generalization ability of learning methods, generation model and discriminant model, classification problem, labeling problem, regression problem.
Assessment: Homework and projects (60%), final exam (40%). |