Abstract
Background
The expectation maximization (EM) algorithm is a common tool for estimating the parameters of Gaussian mixture models (GMM). However, it is highly sensitive to initial value and easily gets trapped in a local optimum.
Method
To address these problems, a new iterative method of EM initialization (MRIPEM) is proposed in this paper. It incorporates the ideas of multiple restarts, iterations and clustering. In particular, the mean vector and covariance matrix of sample are calculated as the initial values of the iteration. Then, the optimal feature vector is selected from the candidate feature vectors by the maximum Mahalanobis distance as a new partition vector for clustering. The parameter values are renewed continuously according to the clustering results.
Results
To verify the applicability of the MRIPEM, we compared it with other two popular initialization methods on simulated and real datasets, respectively. The comparison results of the three stochastic algorithms indicate that MRIPEM algorithm is comparable in relatively high dimensions and high overlaps and significantly better in low dimensions and low overlaps.
Funder
Chinese Universities Scientific Fund
Publisher
Public Library of Science (PLoS)
Reference29 articles.
1. Robust Classification Under ℓ0 Attack for the Gaussian Mixture Model;P Delgosha;SIAM J Math Data Sci,2022
2. Agricultural super green image segmentation method based on Gaussian mixture model combined with Camshift;Jie Jiang;Arabian J Geosci,2021
3. Application of remote sensing image classification based on adaptive Gaussian mixture model in analysis of mountain environment features;N Xu;Arabian J Geosci,2021
4. A study of Gaussian mixture models of color and texture features for image classification and segmentation;H Permuter;Pattern Recognit,2006
5. Using Gaussian Mixtures on Triphone Acoustic Modelling-Based Punjabi Continuous Speech Recognition
Cited by
1 articles.
订阅此论文施引文献
订阅此论文施引文献,注册后可以免费订阅5篇论文的施引文献,订阅后可以查看论文全部施引文献