NDAA09028U Statistical Methods for Machine Learning

The amount and complexity of available data is steadily increasing. To make use of this wealth of information, computing systems are needed that turn the data into knowledge. Machine learning is about developing the required software that automatically analyses data for making predictions, categorizations, and recommendations. Machine learning algorithms are already an integral part of today's computing systems - for example in search engines, recommender systems, or biometrical application. Machine learning provides a set of tools that are widely applicable for data analysis within a diverse set of problem domains such as data mining, search engines, digital image and signal analysis, natural language modeling, bioinformatics, physics, economics, biology, etc.

The purpose of the course is to introduce students to probabilistic data modeling and the most common techniques from statistical machine learning and pattern recognition. The students will obtain a working knowledge of probabilistic data modeling and statistical machine learning for pattern recognition.

This course is relevant for students from among others the studies of Computer Science, E-Science, Bioinformatics, Physics, and Mathematics. 

The course covers the following tentative topic list:

• Foundation of statistical learning, probability theory.
• Likelihood framework, parametric and non-parametric representations. This includes Gaussian distributions, histograms, kernel density estimation, neighborhood based estimation (KNN).
• Classification methods: Linear models, K-Nearest Neighbor (KNN), kernel-based methods such as support vector machines  (SVMs), and neural networks.
• Regression methods: Linear regression, non-linear regression
• K-means clustering and mixture modeling.
• Dimensionality reduction and visualization techniques such as principal component analysis (PCA).