Generating Lambda Term Individuals in Typed Genetic Programming Using Forgetful A*

0435886 - ÚI 2015 RIV US eng C - Konferenční příspěvek (zahraniční konf.)
Křen, T. - Neruda, Roman
Generating Lambda Term Individuals in Typed Genetic Programming Using Forgetful A*.
2014 IEEE Congress on Evolutionary Computation. Piscataway: IEEE CS, 2014, s. 1847-1854. ISBN 978-1-4799-6626-4.
[CEC 2014. IEEE Congress on Evolutionary Computation. Beijing (CN), 06.06.2014-11.06.2014]
Grant CEP: GA ČR GAP202/10/1333; GA MŠMT(CZ) LD13002
Grant ostatní: SVV(CZ) 260 104
Institucionální podpora: RVO:67985807
Klíčová slova: genetic algorithms * geometry * lambda calculus * search problems
Kód oboru RIV: IN - Informatika

Tree based genetic programming (GP) traditionally uses simple S-expressions to represent programs, however more expressive representations, such as lambda calculus, can exhibit better results while being better suited for typed GP. In this paper we present population initialization methods within a framework of GP over simply typed lambda calculus that can be also used in the standard GP approach. Initializations can be parameterized by different search strategies, leading to wide spectrum of methods corresponding to standard ramped halfand- half initialization on one hand, or exhaustive systematic search on the other. A novel geometric strategy is proposed that balances those two approaches. Experiments on well known benchmark problems show that the geometric strategy outperforms the standard generating method in success rate, best fitness value, time consumption and average individual size.
Trvalý link: http://hdl.handle.net/11104/0239718