|Lecture ||Exercise ||Laboratory ||Examination
|2 ||2 ||0 ||Z+Zk
The aim of the course is to introduce students into the field of data mining. Students will be familiarized with methods of data pre-processing, with machine learning algorithms, and with model performance measures and model selection approaches. The gained knowledge will be practised on the real data utilizing open-source software Rapidminer.
- Introduction to data mining. CRISP-DM. Data warehousing. OLAP.
- Pattern recognition - basic concepts. Supervised/unsupervised learning. Classification and regression. Generalization. Overfitting. Bias-variance tradeoff..
- Test set. Cross validation. k-nearest neighbors.
- Cluster analysis.
- Information theory. Decission trees.
- Neural networks I. Threshold neuron. ADALINE. Linear perceptron.
- Neural networks II. Multilayer Perceptron.
- Neural networks III. Radial Basis Function (RBF) Networks.
- Neural networks IV. Self-Organizing Map.
- Support Vector Machines.
- Genetic Algorithms.
- Feature Selection. Feature Extraction.
- Ensemble learning.
- Credit project during the term, finished with oral presentation.
- Witten exam at the end of the term.