Machine Learning and Data mining

The course is divided into two parts. The first part provides a broad introduction to machine learning and statistical pattern recognition. Topics include: supervised learning; unsupervised learning (clustering, dimensionality reduction, kernel methods); learning theory (bias/variance tradeoffs; VC theory; large margins); Neural Networks, Decision Trees, Local Models, Model selection, Combining Multiple Learners, reinforcement learning and adaptive control. The second part of the course applies machine learning techniques to Data Mining concepts and algorithms. It focuses on using tabular data sets and introduces examples of the key algorithmic methods used in each task: Classification, Clustering and Association Rule Induction. It also describes the problems related to mining unstructured data in general and presents selected feature representation and selection methods, information retrieval models, and information extraction approaches. The course will include both hand worked tutorials and applied lab work using the WEKA, Carrot2, and GATE systems.