ETC3250 Introduction to machine learning

Business analytics involves uncovering the hidden information in masses of business data using statistical graphics, models and algorithms. The most widely used prediction and classification models will be covered. Practical skills in applying techniques to different problems will be developed using a suitable software environment that involves doing reproducible analyses. Topics to be covered include dimension reduction with methods such as principal component analysis, supervised learning with methods such as linear models, discriminant analysis, decision trees and forests, support vector machines, neural networks, and unsupervised methods such as k-means clustering. Techniques for numerical optimisation, Monte Carlo simulation, and resampling methods including bootstrap, cross-validation, and bagging will be discussed. Modelling will include nonlinear relationships and nonparametric methods.