An Extremal Optimization Algorithm for Improving Gaussian Mixture Search

Rodica Lung

2022

Abstract

Many standard clustering methods rely on optimizing a maximum likelihood function to reveal internal connections within data. While relying on the same model, alternative approaches may provide better insight into the division of data. This paper presents a new Gaussian mixture clustering approach that uses an extremal optimization algorithm to maximize the silhouette coefficient. The mean and covariance matrix of each component are evolved to maximize each cluster’s priors. Numerical experiments compare the performance of the expectation-maximization algorithm with the new approach on a set of synthetic and real-world data.

Paper Citation

in Harvard Style

Lung R. (2022). An Extremal Optimization Algorithm for Improving Gaussian Mixture Search. In Proceedings of the 14th International Joint Conference on Computational Intelligence (IJCCI 2022) - Volume 1: ECTA; ISBN 978-989-758-611-8, SciTePress, pages 91-96. DOI: 10.5220/0011528000003332

in Bibtex Style

@conference{ecta22,
author={Rodica Lung},
title={An Extremal Optimization Algorithm for Improving Gaussian Mixture Search},
booktitle={Proceedings of the 14th International Joint Conference on Computational Intelligence (IJCCI 2022) - Volume 1: ECTA},
year={2022},
pages={91-96},
publisher={SciTePress},
organization={INSTICC},
doi={10.5220/0011528000003332},
isbn={978-989-758-611-8},
}

in EndNote Style

TY - CONF

JO - Proceedings of the 14th International Joint Conference on Computational Intelligence (IJCCI 2022) - Volume 1: ECTA
TI - An Extremal Optimization Algorithm for Improving Gaussian Mixture Search
SN - 978-989-758-611-8
AU - Lung R.
PY - 2022
SP - 91
EP - 96
DO - 10.5220/0011528000003332
PB - SciTePress