Please use this identifier to cite or link to this item: http://hdl.handle.net/10609/98646
Title: From metaheuristics to learnheuristics: Applications to logistics, finance, and computing
Author: Calvet Liñán, Laura  
Director: Juan, Angel A.  
Abstract: A large number of decision-making processes in strategic sectors such as transport and production involve NP-hard problems, which are frequently characterized by high levels of uncertainty and dynamism. Metaheuristics have become the predominant method for solving challenging optimization problems in reasonable computing times. However, they frequently assume that inputs, objective functions and constraints are deterministic and known in advance. These strong assumptions lead to work on oversimplified problems, and the solutions may demonstrate poor performance when implemented. Simheuristics, in turn, integrate simulation into metaheuristics as a way to naturally solve stochastic problems, and, in a similar fashion, learnheuristics combine statistical learning and metaheuristics to tackle problems in dynamic environments, where inputs may depend on the structure of the solution. The main contributions of this thesis include (i) a design for learnheuristics; (ii) a classification of works that hybridize statistical and machine learning and metaheuristics; and (iii) several applications for the fields of transport, production, finance and computing.
Keywords: metaheuristics
combinatorial optimization
statistics
simheuristics
logistics
Document type: info:eu-repo/semantics/doctoralThesis
Issue Date: 12-Jul-2017
Publication license: http://creativecommons.org/licenses/by-nc-nd/3.0/es/  
Appears in Collections:Tesis doctorals

Files in This Item:
File Description SizeFormat 
tesiCurta.pdfCalvet_Liñán_dissertation2,33 MBAdobe PDFThumbnail
View/Open
Tesi.pdfCalvet_Liñán_dissertation9,1 MBAdobe PDFThumbnail
View/Open