Emphasizing the usefulness ofstandard machinery from applied statistics, this textbook gives an overview of the major applied areas in learning, including coverage of:

¿ classification using standard machinery (naive bayes; nearest neighbor; SVM)

¿ clustering and vector quantization (largely as in PSCS)

¿ PCA (largely as in PSCS)

¿ variants of PCA (NIPALS; latent semantic analysis; canonical correlation analysis)

¿ linear regression (largely as in PSCS)

¿ generalized linear models including logistic regression

¿ model selection with Lasso, elasticnet

¿ robustness and m-estimators

¿ Markov chains and HMM's (largely as in PSCS)

¿ EM in fairly gory detail; long experience teaching this suggests one detailed example is required, which students hate; but once they've been through that, the next one is easy

¿ simple graphical models (in the variational inference section)

¿ classification with neural networks, with a particular emphasis on

¿ autoencoding with neural networks

¿ structure learning