UDK 004.93
INFORMATIVE ATTRIBUTE SELECTION WITH HYBRID SELF-ADJUSTED EVOLUTIONARY OPTIMIZATION ALGORITHM*
S. S. Volkova
Siberian State Aerospace University named after academician M. F. Reshetnev 31, Krasnoyarsky Rabochy Av., Krasnoyarsk, 660014, Russian Federation E-mail: Svetlana.volkova.mail@yandex.ru
An informative attribute selection problem is considered. The problem is solved with the hybrid self-adjusted evolutionary algorithm. The algorithm is used as an optimization method of bandwidth parameters in kernel regression. The algorithm is experimented on the test function with various dimensions. Reliability depends on the dimension function which is also presented. The results of the hybrid self-adjusted algorithm are presented too.
informative attribute selection, kernel regression, genetic algorithm, hybrid self-adjusted genetic algorithm.
References
- Aivazyan, S. A. [еt al.] Applied statistics. Classification and reduction of dimensionality. Moscow, Finansy i statistika, 1989, 607 р.
- Medvedev A. V. Non-parametrical systems of adaptation. Novosibirsk, Nauka, 1983. 174 р.
- Goldberg D. E. Genetic algorithms in search, optimization and machine learning. Reading, MA: Addison-Wesley, 1989.
- Hall P., Li Q. and J. S. Racine, «Nonparametric Estimation of Regression Functions in the Presence of Irrelevant Regressors. Review of Economics and Statistics. 2007. 89. P. 784–789.
- Gomez J. Self-adaptation of operator rates in evolutionary algorithms. Proc. of Genetic and Evolutionary Computation Conference. 2004. Р. 1162–1173.
- Volkova S. Informative attributes selection in non-parametric regression estimation by making use of genetic algorithms. Vestnik Tomskogo gosudarstvennogo universiteta. Ypravlenie, vichislitelnaya technika i informatica. 2013. № 1 (22). P. 40–48.
Volkova Svetlana Sergeevna – Master’s Degree student, Institute of Computer Science and Telecommunications, Siberian State Aerospace University named after academician M. F. Reshetnev. E-mail: Svetlana.volkova.mail@yandex.ru.
