Evolutionary Optimization of Meta Data Metric for Method Recommendation

0425750 - ÚI 2015 RIV US eng C - Conference Paper (international conference)
Kazík, O. - Šmíd, J. - Neruda, Roman
Evolutionary Optimization of Meta Data Metric for Method Recommendation.
Proceedings of the 2013 IEEE Conference on Cybernetics and Intelligent Systems (CIS). Piscataway: IEEE Computer Society, 2013, s. 123-127. ISBN 978-1-4799-1072-4.
[CIS-RAM 2013. IEEE International Conference on Cybernetics and Intelligent Systems /6./ and IEEE International Conference on Robotics, Automation and Mechatronics /6./. Manila (PH), 12.11.2013-15.11.2013]
R&D Projects: GA ČR GAP202/11/1368
Institutional support: RVO:67985807
Keywords : evolutionary algorithms * data mining * metalearning
Subject RIV: IN - Informatics, Computer Science

Metalearning - a method for recommendation the most suitable data-mining algorithm to an unknown dataset - is an important problem that needs to be solved in order to design a completely autonomous data-mining solver. This paper deals with this particular problem by proposing a machinelearning method which recommends the most suitable algorithm to an unknown dataset based on the results of previous datamining experiments. The fundamental idea behind this is that the algorithms will perform similarly on similar datasets. The choice of datasets features - called meta data - is presented and the metric comparing datasets is optimized by means of evolutionary computation.
Permanent Link: http://hdl.handle.net/11104/0232847