@inproceedings{34bcb0da2de845cf84095b3df2983c11,
title = "Harmonious competition learning for gaussian mixtures",
abstract = "This paper proposes a novel automatic model selection algorithm for learning Gaussian mixtures. Unlike EM, we shall further increase the negative entropy of the posterior of latent variables to exert an indirect effect on model selection. The increase of negative entropy can be interpreted as a competition, which corresponds to an annihilation of those components with insufficient data to support. More importantly, this competition only depends on the data itself. Additionally, we seamlessly integrate parameter estimation and model selection into a single algorithm, which can be applied to any kind of parametric mixture model solved by an EM algorithm. Experiments involving Gaussian mixtures show the efficiency of our approach on model selection.",
keywords = "Expectation maximization, Gaussian mixture model, Harmonious competition learning, Model selection",
author = "Liu, \{Guo Jun\} and Tang, \{Xiang Long\}",
year = "2013",
doi = "10.1007/978-3-642-42057-3\_49",
language = "英语",
isbn = "9783642420566",
series = "Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)",
publisher = "Springer Verlag",
pages = "385--392",
booktitle = "Intelligence Science and Big Data Engineering - 4th International Conference, IScIDE 2013, Revised Selected Papers",
address = "德国",
note = "4th International Conference on Intelligence Science and Big Data Engineering, IScIDE 2013 ; Conference date: 31-07-2013 Through 02-08-2013",
}