Machine Learning and Data Analysis

Přednášející:

Filip Železný (gar.)

Cvičící:

Filip Železný (gar.)

Předmět zajišťuje:

katedra kybernetiky

Anotace:

The course explains advanced methods of machine learning helpful for getting insight into data by automatically discovering interpretable data models such as graph- and rule-based. Emphasis is given to the modeling of multi-relational data. The course will also address a theoretical framework explaining why/when the explained algorithms can in principle be expected to work.

Požadavky:

Topics contained in course RPZ.

For details see http://cw.felk.cvut.cz/doku.php/courses/m33sad/start

Osnova přednášek:

1. Cluster analysis, k-means algorithm, hierarchical clustering

2. Principal and independent component analysis.

3. Graphical probabilistic models

4. Grammar and Markov model learning

5. Association rules, the Apriori algorithm

6. Frequent subgraph search

7. Computational learning theory, concept space, PAC learning

8. PAC learning of logic forms

9. Classification rule learning. Algorithms AQ, CN2.

10. Inductive logic programming, least generalization, inverse entailment

11. Learning from logic interpretations, relational decision trees, relational features

12. Statistical relational learning: probablistic relational models

13. Statistical relational learning: Markov logic

14. Learning from texts

Osnova cvičení:

1. Entry test (prerequisite course RPZ). SW tools for machine learning (RapidMiner, WEKA)

2. Data preprocessing, missing and outlying values, clustering

3. Hierarchical clustering, principal component analysis

4. Graphical probabilistic model parameterization

5. Association rule and frequent subgraph search

6. Classification. Learning and ROC curves.

7. Bias vs. variance, ensemble classification

8. Individual task assignment

9. Individual work

10. Individual work

11. Submission of completed assignments

12. Inductive logic programming: the Aleph system

13. Statistical relational learning: the Alchemy system

14. Credits

Cíle studia:

Learn principles of selected methods of data analysis

methods and classifier learning, and elements of learning

theory.

Studijní materiály:

T. Mitchell: Machine Learning, McGraw Hill, 1997

P. Langley: Elements of Machine Learning, Morgan Kaufman 1996

T. Hastie et al: The elements of Statistical Learning, Springer 2001

S. Džeroski, N. Lavrač: Relational Data Mining, Springer 2001

L. Getoor, B. Taskar (eds): Introduction to Statistical Relational

Learning, MIT Press 2007

V. Mařík et al. (eds): Umělá inteligence II, III, IV (Czech)

Poznámka:

Rozsah výuky v kombinované formě studia: 14p+6c

Rozvrh na zimní semestr 2011/2012:

Rozvrh není připraven

Rozvrh na letní semestr 2011/2012:

Rozvrh není připraven

Předmět je součástí následujících studijních plánů:

• Open Informatics - Artificial Intelligence (povinný předmět oboru)
• Open Informatics - Computer Vision and Image Processing (povinný předmět oboru)