Download PDF
Bayesian parameter estimation of discriminative models
Point variable estimates for discriminative models
Using F-tests to compare regression models
Test sets and validation sets
Choosing parametric discriminative probability distributions
Ordinary Least Squares for prediction
Regularising linear regression for prediction
Choosing linear models for prediction
Generalised linear models, the delta rule and binary classification
Generalised linear models and multiclass classification
Classification trees
Regression trees
Bayesian trees
Support Vector Machines (SVMs)
Variational Bayes
The Naive Bayes classifier
The K-Nearest Neighbours (KNN) classifier
Discriminant analysis
Non-parametric regression
Ensemble methods
Ensemble methods for trees
Regularising black box models
Confidence intervals of black box models
Interpreting black box models
Semi-supervised learning
Imputing missing data for prediction
Recommenders
Assume data is mixed gaussian. Use this to estimate classes.
Use LDA with kernel feature spaces.