Population-Based Algorithm Portfolios for Numerical Optimization

In this paper, we consider the scenario that a population-based algorithm is applied to a numerical optimization problem and a solution needs to be presented within a given time budget. Although a wide range of population-based algorithms, such as evolutionary algorithms, particle swarm optimizers,...

Full description

Saved in:
Bibliographic Details
Published in:IEEE transactions on evolutionary computation Vol. 14; no. 5; pp. 782 - 800
Main Authors: Peng, Fei, Tang, Ke, Chen, Guoliang, Yao, Xin
Format: Journal Article
Language:English
Published: New York, NY IEEE 01.10.2010
Institute of Electrical and Electronics Engineers
The Institute of Electrical and Electronics Engineers, Inc. (IEEE)
Subjects:
Algorithm design and analysis
Algorithm portfolios
Algorithmics. Computability. Computer arithmetics
Algorithms
Application software
Applied sciences
Artificial intelligence
Budgeting
Budgets
Computer applications
Computer science
Computer science; control theory; systems
Constituents
Councils
Education
Evolution
Evolutionary algorithms
Evolutionary computation
Exact sciences and technology
global optimization
Laboratories
Mathematical models
metaheuristic algorithms
Migration
numerical optimization
Optimization
Particle swarm optimization
population-based algorithms
Portfolios
Risk
Studies
Theoretical computing
Algorithm portfolios
Script
Probabilistic approach
Evolutionary algorithm
Algorithmics
Global optimum
numerical optimization
Optimization
global optimization
Credit
Heuristic method
Population
Budget
Population dynamics
Swarm intelligence
Metric
Portfolio management
Metamodel
metaheuristic algorithms
population-based algorithms
Mathematical programming
Execution time
ISSN:1089-778X, 1941-0026
Online Access:Get full text
Tags: Add Tag
No Tags, Be the first to tag this record!
Description
Summary:In this paper, we consider the scenario that a population-based algorithm is applied to a numerical optimization problem and a solution needs to be presented within a given time budget. Although a wide range of population-based algorithms, such as evolutionary algorithms, particle swarm optimizers, and differential evolution, have been developed and studied under this scenario, the performance of an algorithm may vary significantly from problem to problem. This implies that there is an inherent risk associated with the selection of algorithms. We propose that, instead of choosing an existing algorithm and investing the entire time budget in it, it would be less risky to distribute the time among multiple different algorithms. A new approach named population-based algorithm portfolio (PAP), which takes multiple algorithms as its constituent algorithms, is proposed based upon this idea. PAP runs each constituent algorithm with a part of the given time budget and encourages interaction among the constituent algorithms with a migration scheme. As a general framework rather than a specific algorithm, PAP is easy to implement and can accommodate any existing population-based search algorithms. In addition, a metric is also proposed to compare the risks of any two algorithms on a problem set. We have comprehensively evaluated PAP via investigating 11 instantiations of it on 27 benchmark functions. Empirical results have shown that PAP outperforms its constituent algorithms in terms of solution quality, risk, and probability of finding the global optimum. Further analyses have revealed that the advantages of PAP are mostly credited to the synergy between constituent algorithms, which should complement each other either over a set of problems, or during different stages of an optimization process.
Bibliography:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
ObjectType-Article-2
ObjectType-Feature-1
content type line 23
ISSN:1089-778X
1941-0026
DOI:10.1109/TEVC.2010.2040183