Survey of Statistical Machine Learning Concepts (part 1)

This two-part workshop series led by Dr. Duncan Temple Lang aims to provide an overview of relatively modern statistical machine learning concepts and methods including how they have evolved and what problems they address. At the end of the workshops participants should have improved familiarity with the various methods and be able to articulate their applications and critique their use. This workshop will uncover some of the math behind the techniques, but will emphasize the heuristic concepts. The topics range from classification vs prediction vs inference, parametric vs non-parametric, supervized/unsupervized/semisupervized learning, bias-variance, overfitting, regularization, dimension reduction, cross-validation, ensembles, Bayesian versus frequentist methods, feature engineering, and other specific methods.

Prerequisites: This is a largely conceptual workshop, no specific programming skills are required.

Resources:

Slides & Examples
Recording