V. Linear classification
Given by Sven Laur
Brief summary: General description of linear classifiers. Components of the linear discriminant functions: weight vector and bias term. Fisher discriminant, Least-squares methods, Perceptron, Batch and online training, Various cost functions used in training and resulting training algorithms.
Slides: PDF PDF(modified) Δ PDF(2015)
Auxiliary materials: PDF
Video: Video Screencast(2015)
Literature:
- Duda, Hart & Stork: Patter Classification: Linear discriminant functions.