1. Overview of ML methods
2. What is learning and relation between learning and intelligence
3. Overview of necessary background
4. Advanced attribute evaluation measures
5. Advanced methods for estimating performance
6. Combining ML algorithms
7. Bayesian learning
8. Calibration of probabilities, Explanation of individual predictions
9. Numerical ML methods
10. Artificial neural networks: Hopfield NN, RBF, Deep NN
11. Unsupervised learning: clustering, Association rules, spatial DM
12. Constructive induction, reliability of predictions
13. Text mining, Matrix factorization, Arcehtypal analysis
14. Other approaches to ML
15. Introduction to formal learning theory
Practical applications of the knowledge gained through lectures. The emphasis is on the autonomous work of students with the help of assistants. Students will, in small groups, independently solve real-life problems under the supervision of different experts in ML and DM. The groups will describe their solutions in written reports and present them in short presentations and through those will receive their mark from lab. work.
Igor Kononenko and Matja.ž Kukar: Machine Learning and Data Mining. Horwood Publ., 2007.
David J. Hand, Heikki Mannila, Padhraic Smyth: Principles of Data Mining. The MIT Press, 2001.
Ian H. Witten, Eibe Frank: Data Mining: Practical Machine Learning Tools and Techniques. Morgan Kaufmann, 1999.
The goal is to present the basics and the basic principles of machine learning (ML) methods, the basic ML algorithms and their usage in practice for knowledge discovery from data, data mining (DM) and for learning classification and regression models. Students will practically apply the theoretical knowledge on real problems from scientific and business environment. The students shall be able to decide for a given problem which of the presented techniques should be used, and to develop a prototype solution.
Competences in computer and information science granting access to further study at 3rd cycle doctoral programmes. The ability to transmit knowledge to co-workers in technology and research groups. The ability to understand and apply computer and information science knowledge to other technical and relevant fields (economics, organisational science, etc), The ability to apply acquired knowledge in independent work for solving technical and scientific problems in computer and information science, the ability to upgrade acquired knowledge. The ability to search knowledge sources and to search for resources and critically evaluate information.
Developing skills in critical, analytical and synthetic thinking.
With successful completion of this course the student will:
- be able to use the expertise of several techniques and methods, used for data modelling with ML, for analysis, synthesis and anticipation of solutions and their consequences for target problems using the scientific methodology.
- be able to use of the presented methods on target problems from scientific and business environment. Will understand and use the tools for modelling and data mining. (S)he will flexibly use the knowledge in practice
- be able to bind together the knowledge from different fields to apply it in practice.
- differentiate among different approaches to machine learning, their advatages and disadvantages and wil be able to select the appropriate method for solving particular data mining problem
-be able, using the principles of models, learned from data, to improve the usability and the performance of the analysed system.
Lectures, exercises with oral demonstrations and presentations, seminar works and solving of home-works, which stimulate online learning. The emphasis is on an online study and an independent exercises and seminars.
Continuing (homework, midterm exams, project work)
Final (written and oral exam)
grading: 5 (fail), 6-10 (pass) (according to the Statute of UL)
Pet najpomembnejših del:
I.Kononenko, M.Kukar: Introduction to Machine Learning and Data Mining: Introduction to Principles and Algorithms, Horwood Publishing, 2007. XIX, 454 pages.
E. Štrumbelj, I.Kononenko. An efficient explanation of individual classifications using game theory. Journal of machine learning research, ISSN 1532-4435, 2010, vol. 11, no. , p. 1-18
Z. Bosnić, I. Kononenko. Automatic selection of reliability estimates for individual regression predictions. Knowledge engineering review, ISSN 0269-8889, 2010, vol. 25, no. 1, p. 27-47.
Robnik-Šikonja, M., Kononenko, I. Theoretical and empirical analysis of ReliefF and RReliefF. Machine Learning. [Print ed.], 2003, vol. 53, str. 23-69.
Machine learning for medical diagnosis: History, state of the art and perspective, Invited paper, Artificial Intelligence in Medicine - ISSN 0933-3657, 23:89-109, 2001.
Celotna bibliografija je dostopna na SICRISu: