Adaptive Online Time Allocation to Search Algorithms
Abstract
Given is a search problem or a sequence of search problems, as well as a set of potentially useful search algorithms. We propose a general framework for online allocation of computation time to search algorithms based on experience with their performance so far. In an example instantiation, we use simple linear extrapolation of performance for allocating time to various simultaneously running genetic algorithms characterized by different parameter values. Despite the large number of searchers tested in parallel, on various tasks this rather general approach compares favorably to a more specialized state-of-the-art heuristic; in one case it is nearly two orders of magnitude faster.
Cite
Text
Gagliolo et al. "Adaptive Online Time Allocation to Search Algorithms." European Conference on Machine Learning, 2004. doi:10.1007/978-3-540-30115-8_15Markdown
[Gagliolo et al. "Adaptive Online Time Allocation to Search Algorithms." European Conference on Machine Learning, 2004.](https://mlanthology.org/ecmlpkdd/2004/gagliolo2004ecml-adaptive/) doi:10.1007/978-3-540-30115-8_15BibTeX
@inproceedings{gagliolo2004ecml-adaptive,
title = {{Adaptive Online Time Allocation to Search Algorithms}},
author = {Gagliolo, Matteo and Zhumatiy, Viktor and Schmidhuber, Jürgen},
booktitle = {European Conference on Machine Learning},
year = {2004},
pages = {134-143},
doi = {10.1007/978-3-540-30115-8_15},
url = {https://mlanthology.org/ecmlpkdd/2004/gagliolo2004ecml-adaptive/}
}