Module Title:
Machine Learning and Computational Statistics
Module Code:
DSC6135

Module Content

• Statistical learning theory framework, stochastic gradient descent,
• Matrix/vector differentiation
• Excess risk decomposition, L1/L2 regularisation, Lasso algorithms, sub-gradient descent
• Loss functions, convex optimization, Support Vector Machine
• ernels, kernel ridge regression, kernelised Support Vector Machine
• Trees, bias and variance decomposition
• nsemble methods: bootstrap, bagging, random forest, AdaBoost
• Gradient boosting, neural networks Spring Break
• Natural exponential families and generalised linear models
• Bayesian networks, class-conditional models, naïve Bayes
• Clustering, Gaussian mixture models, EM algorithm
• Bayesian methods, hierarchical models, Gibbs sampling, Singular Value Decomposition, PCA, Linear Discriminant Analysis