This semester, we are reading the following book: Introduction to Machine Learning (by Alpaydin). Our weekly meetings are held on Fridays 12:30-2pm in Room 115, Building 63.
A good news for HDR students: According to Sue McKemmish (postgraduate coordinator), you can claim hours for active participation in the reading group as part of your HDR-training hours.
In the first half of each meeting, a volunteer presents the assigned chapter, and in the second half, people start discussing the chapter. Based on the responses so far, the following table shows the assignment of people to chapters. If you haven't responded yet, please let me know your preferences among the chapters that have not been assigned yet.
Summary of Covered Chapters
Chapter 2 (Supervised Learning): Hypothesis Class, Version Space, Vapnik–Chervonenkis (VC) Dimension, Probably Approximately Correct (PAC) Learning, Generalization Error, Model Selection, Cross Validation
Chapter 3 (Bayesian Decision Theory): Bayes Rule, Losses and Risks, Discriminant Functions, Utility Theory, Frequentist vs Bayesian Statistics
Chapter 4 (Parametric Methods): Parameter Estimation (Maximum Likelihood, Maximum a Posteriori, Bayes’ Estimator), Bias and Variance (consistency) of Estimators,
Parametric Density Estimation/Classification/Regression, Bias and Variance of the Estimator for Squared-Error Loss (in Regression), Interplay among Bias/Variance/Model Complexity/Overfitting/Underfitting
Chapter 5 (Multivariate Methods): Multivariate Data/Parameters, Estimation of Missing Values, Multivariate Normal Distribution, Naive Bayes, Multivariate Classification/Regression
Chapter 6 (Dimensionality Reduction): PCA, Factor Analysis, Multidimensional Scaling, Linear Discriminant Analysis, Isomap, Locally Linear Embedding
Chapter 7 (Clustering): Density Estimation, K-Means, Expectation-Maximization (EM), Mixtures of Latent Variable Models, Hierarchical Clustering, Agglomerative Clustering, Choosing "k"
Chapter 8 (Nonparametric Methods): Kernel Density Estimation, k-Nearest Neighbor Estimator, Nonparametric Regression, Choosing the Bandwidth
Chapter 9 (Decision Trees): Classification Trees, Regression Trees, Multivariate Trees
Chapter 10 (Linear Discrimination): Likelihood- vs. Discriminant-based
Classification, Generalized Linear Model, From Discriminants to Posteriors, Logistic Discrimination, Discrimination by Regression
Chapter 11 (Multilayer Perceptrons): Neural Networks, Training a Perceptron: Regression, Multilayer Perceptrons, Backpropagation, Two-Class Discrimination, Multiple Hidden Layers, Structured MLP, Weight Sharing, Bayesian Learning, Dimensionality Reduction, Time-Delay Neural Networks, Recurrent Networks
Tutorial (Non-parametric Bayesian Models): species samplers as simple mixture models, Chinese restaurant process (CRP), Pitman-Yor process, hierarchical Pitman-Yor process, challenge problems
|
|