1. komponenta

Lecture type	Total
Lectures	45

* Load is given in academic hour (1 academic hour = 45 minutes)

COURSE AIMS AND OBJECTIVES: Introducing students to various data mining approaches. Pointing out advantages, disadvantages of each approach, explaining input data format and the corresponding output results.

COURSE DESCRIPTION AND SYLLABUS:
1. Introduction. About data mining and input tabular data. (1 week)
2. Basic measures from data mining and information theory. Repetition of basic measures from statistics (variance, expectation etc.). Definition of entropy, information gain, mutual information. Basic measures to evaluate a model (frequently used in data mining) accuracy, precision etc. (1 week)
3. Introduction to decision trees. Description of a decision tree model, training of this model, making predictions. Evaluation of a decision tree model. (1 week)
4. Rule mining. Description of rule sets, training, goal, evaluation. (1 week)
5. Closed and frequent itemsets. Algorithms for closed and frequent itemset mining, goals, type of knowledge provided. (1 week)
6. Association rule mining. Problem description, description of algorithm for association rule mining and the obtainable knowledge. Basic applications with examples. (1 week)
7. Subgroup discovery. Problem description, algorithm for finding subgroups, evaluation and applications. (1 week)
8. Exceptional model mining. Problem description, algorithm for finding exceptional models, analyses of obtained knowledge. (1 week)
9. Conceptual clustering. Basics of clustering. Description of conceptual clustering emphasising difference from basic clustering, basic algorithm for conceptual clustering, analyses of obtainable knowledge. (2 weeks)
10. Redescription mining. Description of a problem, the most common approaches and analyses of obtainable knowledge. (1 week)
11. Basic post-processing and data visualization. Analyses of models, rule sets, visualization of input data and the results. (2 weeks)

1. Data mining and knowledge discovery handbook, Lior Rokach, Springer, New York, 2005.
2. Foundations of Rule Learning, Johannes Furnkranz, Dragan Gamberger, Nada Lavrač, Springer Science & Business Media, 2012.
3. Redescription mining, Esther Galbrun, Pauli Miettinen, Springer, Cham, 2017.
4. Data mining: the textbook, Aggarwal, Springer, New York, 2015.
5. Journal of Machine learning research, Nada Lavrač et al. Subgroup Discovery with CN2-SD, 2004.
6. Artificial Intelligence Reviews, A review of conceptual clustering algorithms, 2019.

Izborni predmet 3 - Regular study - Financial and Business Mathematics

Izborni predmet 4, 5, 6 - Regular study - Financial and Business Mathematics

Izborni predmet 4, 5, 6 - Regular study - Financial and Business Mathematics