A new iterative initialization of EM algorithm for Gaussian mixture models

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. To address these problems, a new iterative method of EM initialization (MRIPEM) is...

Celý popis

Uloženo v:
Podrobná bibliografie
Vydáno v:PloS one Ročník 18; číslo 4; s. e0284114
Hlavní autoři: You, Jie, Li, Zhaoxuan, Du, Junli
Médium: Journal Article
Jazyk:angličtina
Vydáno: United States Public Library of Science 13.04.2023
Public Library of Science (PLoS)
Témata:
Algorithms
Analysis
Cluster Analysis
Clustering
Computer and Information Sciences
Covariance matrix
Data mining
Eigenvectors
Evaluation
Gaussian processes
Iterative methods
Learning
Medicine and Health Sciences
Methods
Mixtures
Normal Distribution
Optimization
Parameter estimation
Parameter sensitivity
People and Places
Physical Sciences
Probabilistic models
Probability
Probability distribution
Random variables
Research and Analysis Methods
Statistical analysis
Stochasticity
China
ISSN:1932-6203, 1932-6203
On-line přístup:Získat plný text
Tagy: Přidat tag
Žádné tagy, Buďte první, kdo vytvoří štítek k tomuto záznamu!
Popis
Shrnutí: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. 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. 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.
Bibliografie:ObjectType-Article-1
SourceType-Scholarly Journals-1
ObjectType-Feature-2
content type line 14
content type line 23
Competing Interests: The authors have declared that no competing interests exist.
ISSN:1932-6203
1932-6203
DOI:10.1371/journal.pone.0284114