Adaptive and Multilevel Metaheuristics by Konstantin Chakhlevitch, Peter Cowling (auth.), Carlos

By Konstantin Chakhlevitch, Peter Cowling (auth.), Carlos Cotta, Marc Sevaux, Kenneth Sörensen (eds.)

One of the keystones in sensible metaheuristic problem-solving is the truth that tuning the optimization strategy to the matter into consideration is essential for reaching best functionality. This tuning/customization is mostly within the arms of the set of rules dressmaker, and regardless of a few methodological makes an attempt, it principally continues to be a systematic artwork. shifting part of this customization attempt to the set of rules itself -endowing it with clever mechanisms to self-adapt to the matter- has been a protracted pursued target within the box of metaheuristics.

These mechanisms can contain varied points of the set of rules, corresponding to for instance, self-adjusting the parameters, self-adapting the functioning of inner elements, evolving seek options, etc.

Recently, the assumption of hyperheuristics, i.e., utilizing a metaheuristic layer for adapting the hunt by means of selectively utilizing assorted low-level heuristics, has additionally been becoming more popular. This quantity provides fresh advances within the sector of adaptativeness in metaheuristic optimization, together with up to date reports of hyperheuristics and self-adaptation in evolutionary algorithms, in addition to innovative works on adaptive, self-adaptive and multilevel metaheuristics, with program to either combinatorial and non-stop optimization.

Show description

Read or Download Adaptive and Multilevel Metaheuristics PDF

Best nonfiction_5 books

Basic Coastal Engineering: 3rd (Third) edition

Simple Coastal Engineering, third variation, deals the fundamentals on monochromatic and spectral surgace wave mechanics, coaster water point adaptations, coastal constructions and coastal sedimentary strategies. It additionally offers the mandatory historical past from which the reader can pursue a extra complex strudy of a number of the theroretical and utilized features of coastal hydrodynamics and layout.

TI-Nspire For Dummies (For Dummies (Computer Tech))

The up to date advisor to the latest graphing calculator from Texas InstrumentsThe TI-Nspire graphing calculator is renowned between highschool and faculty scholars as a beneficial software for calculus, AP calculus, and college-level algebra classes. Its use is permitted at the significant collage front tests. This e-book is a nuts-and-bolts consultant to operating with the TI-Nspire, offering every little thing you must wake up and operating and supporting you get the main out of this high-powered math instrument.

Extra resources for Adaptive and Multilevel Metaheuristics

Sample text

Annals of Operations Research 92, 363–380 (1999) 44. : Reinforcement learning: a survey. Journal of Artificial Intelligence Research 4, 237–285 (1996) 45. : Channel assignment in cellular communication using a Great Deluge hyper-heuristic. In: Proceedings of the 2004 IEEE International Conference on Networks (ICON 2004), Singapore, November 16-19 (2004) 46. : Channel assignment optimisation using a hyperheuristic. In: Proceedings of the 2004 IEEE Conference on Cybernetics and Intelligent Systems (CIS 2004), Singapore, December 1-3 (2004) 47.

Many successful applications have been reported, particularly in the sub-field of Evolution Strategies for problems in the continuous domain. In this chapter we examine the motivation and features necessary for successful self-adaptive learning to occur. Since a number of works have dealt with the continuous domain, this chapter focusses particularly on its aspects that arise when it is applied to combinatorial problems. We describe how self-adaptation may be use to control not only the parameters defining crossover and mutation, but also how it may be used to control the very definition of local search operators used within hybrid evolutionary algorithms (so-called memetic algorithms).

Xn , σ . Mutations are then realised by replacing x1 , . . , xn , σ by x1 , . . , xn , σ . E. Smith distribution with mean 0 and standard deviation τ . Since N (0, τ ) = τ · N (0, 1), the full mutation mechanism is: σ = σ · eτ ·N (0,1), (1) xi = xi + σ · Ni (0, 1). (2) In these formulas N (0, 1) denotes a draw from the standard normal distribution, and since step-sizes very close to zero will have on average a negligible effect, a boundary rule is then applied to maintain σ at or above a small threshold value.

Download PDF sample

Rated 4.20 of 5 – based on 30 votes