An Agent-based Approach to Decentralized Global Optimization - Adapting COHDA to Coordinate Descent

Joerg Bremer, Sebastian Lehnhoff

2017

Abstract

Heuristics like evolution strategies have been successfully applied to optimization problems with rugged, multi-modal fitness landscapes, to non-linear problems, and to derivative free optimization. Parallelization for acceleration often involves domain specific knowledge for data domain partition or functional or algorithmic decomposition. We present an agent-based approach for a fully decentralized global optimization algorithm without specific decomposition needs. The approach extends the ideas of coordinate descent to a gossiping like decentralized agent approach with the advantage of escaping local optima by replacing the line search with a full 1-dimensional optimization and by asynchronously searching different parts of the search space using agents. We compare the new approach with the established covariance matrix adaption evolution strategy and demonstrate the competitiveness of the decentralized approach even compared to a centralized algorithm with full information access. The evaluation is done using a bunch of well-known benchmark functions.

Paper Citation

in Harvard Style

Bremer J. and Lehnhoff S. (2017). An Agent-based Approach to Decentralized Global Optimization - Adapting COHDA to Coordinate Descent . In Proceedings of the 9th International Conference on Agents and Artificial Intelligence - Volume 1: ICAART, ISBN 978-989-758-219-6, pages 129-136. DOI: 10.5220/0006116101290136

in Bibtex Style

@conference{icaart17,
author={Joerg Bremer and Sebastian Lehnhoff},
title={An Agent-based Approach to Decentralized Global Optimization - Adapting COHDA to Coordinate Descent},
booktitle={Proceedings of the 9th International Conference on Agents and Artificial Intelligence - Volume 1: ICAART,},
year={2017},
pages={129-136},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0006116101290136},
isbn={978-989-758-219-6},
}

in EndNote Style

TY - CONF
JO - Proceedings of the 9th International Conference on Agents and Artificial Intelligence - Volume 1: ICAART,
TI - An Agent-based Approach to Decentralized Global Optimization - Adapting COHDA to Coordinate Descent
SN - 978-989-758-219-6
AU - Bremer J.
AU - Lehnhoff S.
PY - 2017
SP - 129
EP - 136
DO - 10.5220/0006116101290136