Cipher: 2206
Nomenclature: IT methods of knowledge discovery
Study programme: Molecular biosciences
Module: Bioinformatics
Case holder:

Izv.prof.dr.sc. Anita Papić
Institution of the case holder:

Faculty of Humanities and Social Sciences, University of Josip Juraj Strossmayer in Osijek
Contributors - Contractors:
Subject status: Electoral College
The year in which the case is submitted: Year I
The semester in which the case is submitted: Semester II
Subject objective:

Learn methods of discovering knowledge based on inductive machine learning techniques and understand their application in medicine, genetics and chemistry.
Case contents:

Databases and the need for methods of their analysis. Analysis of data with the aim of predicting and classifying unclassified examples. Discovering knowledge in scientific research work with application in the formation of new knowledge and the direction of research. The process of analyzing data using artificial intelligence methods. Algorithms of inductive learning, association learning, subgroup detection, detection of exceptions and errors. Data clusters. Induction from data timelines. Visualization of discovered knowledge. Learning from relational databases. Comparison of knowledge extracted using various systems and statistical methods. Application of statistical methods in verification and detection of confirming factors. Practical work on real medical, bioinformatic and chemical problems with an emphasis on data selection, pre-processing and transformation of data, knowledge generation in the form of rules and their expert interpretation. Systems used: Data Mining Server, Weka and Tanagra.
Learning outcomes: competences, knowledge, skills that the subject develops:

1. Use knowledge discovery methods on their and publicly available data in scientific and development research.
2. Analyse your own data and data from scientific and developmental research.
3. Valorise data for the purpose of solving actual medical, bioinformatic and chemical problems.
4. Formulate hypotheses suitable for theoretical and experimental verification and scientific publications.
ECTS Credits 6
Lectures 20
Seminars (IS) 5
Exercises (E) 5
Altogether 30
The way of teaching and acquiring knowledge:

Mandatory independent work, preferably on your own data.
Ways of teaching and acquiring knowledge: (notes)
Monitoring and evaluating students (mark in fat printing only relevant categories) Attendance, Mandatory seminar work
Rating method: Oral exam, Project
Mandatory literature:

M.Berthold, D.J:Hand: Intelligent Data Analysis – An Introduction. Springer in 1999.
H.Liu, H.Motoda: Instance Selection and Construction for Data Mining. Kluwer 2001
D.Mladenić, N.Lavrač, M.Bohanec, S.Moyle: Data Mining and Decision Support –Integration and Collaboration. Kluwer 2003
S.Jerryski, N.Lavrač: Relational Data Mining. Springer in 2001.
A.A.Freitas: Data Mining and Knowledge Discovery with Evolutionary Algorithms. Springer 2002.
Supplementary (recommended) literature:

L.DeRaedt: Advances in Inductive Logic Programming. 1995 IOS Press
How to monitor the quality and performance performance (evaluation):

The success of the course will be evaluated annually by the joint expert committee of the Ruđer Boskovic Institute, the University of Dubrovnik and the University of Josip Juraj Strossmayer in Osijek based on exam success and surveys.