Data Mining II


1 - Make predictions from data. 2 - Know the main problems related to predictions based on data ("data driven") 3 - Know the main techniques: 3.1 - Classical methods: regression, interpolation, extrapolation 3.2 - Bayesian Decisions 3.3 - 3.4 instances based Systems - Decision trees 3.6 - 3.6 Neural networks - Ensambles

General characterization





Responsible teacher


Weekly - Available soon

Total - Available soon

Teaching language

Portuguese. If there are Erasmus students, classes will be taught in English




Mitchell, T., (1997) ¿Machine Learning¿, McGraw Hill.; Berry, M.J.A. and G.S. Linoff, ¿Data Mining Techniques; for marketing, sales and customer support¿. 1997, John Wiley & Sons.; Hand, D. J., Mannila, H., Smyth, P. (2001) ¿Principles of Data Mining (Adaptive Computation and Machine Learning)¿, MIT Press; 0; 0

Teaching method

The course is mainly based on lecture and practical classes. The practical sessions include exposure of concepts and methodologies, sample resolution, discussion and interpretation of results. Practical work, which is very significant in this course is done by students outside the classroom, but is evaluated.

Evaluation method

The evaluation is done through homework (20% of final grade), a practical group (20%), and one final exam (60%)

Subject matter

The course is organized in seven Learning Units (UA):
UA1. Introduction to forecasting methods in Data Mining
UA2. Data pre-processing and error estimates
UA3. Decision theory and Bayesian
UA4 Learning and classification systems
UA5 Decision trees
UA6. Neural Networks
UA7. Ensambles