Williams, H. and Bishop, Mark (J. M.). 2014. Stochastic Diffusion Search: A Comparison of Swarm Intelligence Parameter Estimation Algorithms with RANSAC. Algorithms, 7(2), pp. 206-228. ISSN 1999-4893 [Article]

Preview

Text (Stochastic Diffusion Search: A Comparison of Swarm Intelligence Parameter Estimation Algorithms with RANSAC) algorithms-07-00206-v3.pdf Available under License Creative Commons Attribution. Download (324kB) | Preview

Abstract or Description

Stochastic diffusion search (SDS) is a multi-agent global optimisation technique based on the behaviour of ants, rooted in the partial evaluation of an objective function and direct communication between agents. Standard SDS, the fundamental algorithm at work in all SDS processes, is presented here. Parameter estimation is the task of suitably fitting a model to given data; some form of parameter estimation is a key element of many computer vision processes. Here, the task of hyperplane estimation in many dimensions is investigated. Following RANSAC (random sample consensus), a widely used optimisation technique and a standard technique for many parameter estimation problems, increasingly sophisticated data-driven forms of SDS are developed. The performance of these SDS algorithms and RANSAC is analysed and compared for a hyperplane estimation task. SDS is shown to perform similarly to RANSAC, with potential for tuning to particular search problems for improved results.

Item Type:	Article
Identification Number (DOI):	https://doi.org/10.3390/a7020206
Keywords:	optimisation; search; swarm; intelligence; stochastic; diffusion; RANSAC; hyperplane; estimation
Departments, Centres and Research Units:	Computing
Dates:	Date Event 2014 Published
Item ID:	10865
Date Deposited:	05 Nov 2014 09:51
Last Modified:	29 Apr 2020 16:03
Peer Reviewed:	Yes, this version has been peer-reviewed.
URI:	https://research.gold.ac.uk/id/eprint/10865

View statistics for this item...

Edit Record (login required)