STA314H5 • Introduction to Statistical Learning

Description

A thorough introduction to the basic ideas in supervised statistical learning with a focus on regression and a brief introduction to classification. Methods covered will include multiple linear regression and its extensions, k-nn regression, variable selection and regularization via AIC,BIC, Ridge and lasso penalties, non-parametric methods including basis expansions, local regression and splines, generalized additive models, tree-based methods, bagging, boosting and random forests. Content will be discussed from a statistical angle, putting emphasis on uncertainty quantification and the impact of randomness in the data on the outcome of any learning procedure. A detailed discussion of the main statistical ideas behind crossvalidation, sample splitting and re-sampling methods will be given. Throughout the course, R will be used as software, a brief introduction will be given in the beginning.

Prerequisites
Corequisites
Exclusions
Enrolment Limits
Priority is given to students enrolled in Statistics Specialist or Major programs.
Distribution Requirement
Science
Total Instructional Hours
36L/12T
Mode of Delivery
In Class
Program Area
Statistics, Applied