Boolean Factor Analysis by Expectation-Maximization Method

Authors: A. A. Frolov, Dušan Húsek, P.Y. Polyakov
Publication year: 2013
Publication type: Konferenční příspěvek (mezinárodní konf.)
Periodical: Proceedings of the Third International Conference on Intelligent Human Computer Interaction IHCI 2011
Projects: Decompositions of matrices with binary and ordinal data: theory, algorithms, and complexity
Methods of Artificial Inteligence in GIS
Bibliographic citation: FROLOV, A. A., HÚSEK, Dušan, POLYAKOV, P.Y. Boolean Factor Analysis by Expectation-Maximization Method. In: KUDĚLKA, M., POKORNÝ, J., SNÁŠEL, V., ABRAHAM, A., eds. Proceedings of the Third International Conference on Intelligent Human Computer Interaction IHCI 2011. Heidelberg: Springer, 2013, S. 243-254. Advances in Intelligent Systems and Computing, 179. ISBN 978-3-642-31602-9. ISSN 2194-5357
Dostupné z: doi:10.1007/978-3-642-31603-6_21
DOI: 10.1007/978-3-642-31603-6_21
ID: 368469

Abstract

Boolean factor analysis is one of the most efficient methods to reveal and to overcome informational redundancy of high-dimensional binary signals. In the present study, we introduce new Expectation-Maximization method which maximizes the likelihood of Boolean factor analysis solution. Using the so-called bars problem benchmark, we compare efficiencies of the proposed method with Dendritic Inhibition neural network.