Introduction to data mining

The course will in theory and through practical exercises and hands-on lectures include the following topics:
Introduction to business intelligence. Typical applications. Role of information technology. Knowledge-based systems.
Computer-assisted decision support. Decision support models. Treatment of uncertain and incomplete data. Explanation and analysis.
Methods and techniques for group decision making.
Introduction to techniques of data mining and knowledge discovery in data bases, with emphasis on their application in business intelligence. Data preprocessing, modelling. Supervised and unsupervised learning.
Data and model visualization.
Data clustering.
Business intelligence on the world-wide-web. Page ranking. Analysis of social networks.
Recommendation systems.
Data analysis toolboxes for business intelligence and their integration in information systems. Interface design of decision support systems.
Psychosocilogical and ethical issues.

Tan, P.-N., Steinbach, M., and Kumar, V. (2006) Introduction to Data Mining, Pearson Education.
Segaran, T. (2007) Programming Collective Intelligence, O'Reilly.
Dokumentacija prosto dostopnih programov za podatkovno analitiko (Orange, na strani http://orange.biolab.si, scikit-learn na strani http://scikit-learn.org in numpy na strani http://www.numpy.org).

The aim of this course is an introduction to business intelligent methods and tools that were developed within computer science. Students will learn how to identify potential applications of business intelligence in practice. During the course, they will apply their methodological and development knowledge on real-life applications. In particular, the course will focus on data clustering, recommendations systems, association rule mining, inference of predictive models from structured and textual data, and on decision support techniques.

Knowledge and understanding: Familiarity and practical understanding of business intelligence techniques.
Application: Utility of business intelligence approaches in information systems and on the web.
Reflection: Competence to determine where and when utility of business intelligence can provide competitive gains. Ability to identify the most useful techniques for a given practical problem.
Transferable skills: Programming in Python. Data mining. Cognitive aspects of decision-making.

Lectures using modern audio-visual equipment. Individual and group-based project assignments. Emphasis on practical exercises.

Continuing (homework and laboratory exercises)
Final (written and oral exam)
grading: 5 (fail), 6-10 (pass) (according to the Statute of UL)

TOPLAK, Marko, MOČNIK, Rok, POLAJNAR, Matija, BOSNIĆ, Zoran, CARLSSON, Lars, HASSELGREN, Catrin, DEMŠAR, Janez, BOYER, Scott, ZUPAN, Blaž, STÅLRING, Jonna. Assessment of machine learning reliability methods for quantifying the applicability domain of QSAR regression models. Journal of chemical information and modeling, ISSN 1549-9596. [Print ed.], Feb. 2014, vol. 54, no. 2, str. 431-441, graf. prikazi. [COBISS-SI-ID 10466388]
ŽITNIK, Marinka, JANJIĆ, Vuk, LARMINIE, Chris, ZUPAN, Blaž, PRŽULJ, Nataša. Discovering disease-disease associations by fusing systems-level molecular data. Scientific reports, ISSN 2045-2322, 2013, str. 1-9, ilustr. [COBISS-SI-ID 10253396]
DEMŠAR, Janez, CURK, Tomaž, ERJAVEC, Aleš, GORUP, Črtomir, HOČEVAR, Tomaž, MILUTINOVIĆ, Mitar, MOŽINA, Martin, POLAJNAR, Matija, TOPLAK, Marko, STARIČ, Anže, ŠTAJDOHAR, Miha, UMEK, Lan, ŽAGAR, Lan, ŽBONTAR, Jure, ŽITNIK, Marinka, ZUPAN, Blaž. Orange : data mining toolbox in Python. Journal of machine learning research, ISSN 1532-4435. [Print ed.], Aug. 2013, vol. 14, str. 2349-2353. [COBISS-SI-ID 10118740]
ŽITNIK, Marinka, ZUPAN, Blaž. NIMFA : a Python library for nonnegative matrix factorization. Journal of machine learning research, ISSN 1532-4435. [Print ed.], Mar. 2012, vol. 13, str. 849-853. [COBISS-SI-ID 9067604]