Szczegóły publikacji
Opis bibliograficzny
Two-phase strategy managing insensitivity in global optimization / Jakub SAWICKI, Maciej SMOŁKA, Marcin ŁOŚ, Robert SCHAEFER, Piotr FALISZEWSKI // W: Applications of evolutionary computation : 20th European conference, EvoApplications 2017 : Amsterdam, The Netherlands, April 19–21, 2017 : proceedings, Pt. 1 / eds. Giovanni Squillero, [et al.]. — Switzerland : Springer International Publishing, cop. 2017. — (Lecture Notes in Computer Science ; ISSN 0302-9743 ; LNCS 10199). — ISBN: 978-3-319-55848-6; e-ISBN: 978-3-319-55849-3. — S. 266–281. — Bibliogr. s. 280–281, Abstr. — Publikacja dostępna online od: 2017-03-25
Autorzy (5)
Słowa kluczowe
Dane bibliometryczne
ID BaDAP | 105205 |
---|---|
Data dodania do BaDAP | 2017-05-15 |
DOI | 10.1007/978-3-319-55849-3_18 |
Rok publikacji | 2017 |
Typ publikacji | materiały konferencyjne (aut.) |
Otwarty dostęp | |
Wydawca | Springer |
Konferencja | 20th European conference on Applications of Evolutionary Computation |
Czasopisma/serie | Lecture Notes in Computer Science, Theoretical Computer Science and General Issues |
Abstract
Solving ill-posed continuous, global optimization problems remains challenging. For example, there are no well-established methods for handling objective insensitivity in the neighborhood of solutions, which appears in many important applications, e.g., in non-invasive tumor tissue diagnosis or geophysical exploration. The paper presents a complex metaheuristic that identifies regions of objective function’s insensitivity (plateaus). The strategy is composed of a multi-deme hierarchic memetic strategy coupled with random sample clustering, cluster integration, and special kind of multiwinner selection that allows to breed the demes and cover each plateau separately. We test the method on benchmarks with multiple non-convex plateaus and evaluate how well the plateaus are covered.