Learning mixture models using a genetic version of the EM algorithm

Aleix M. Martínez, Jordi Vitrià

    Research output: Contribution to journalArticleResearchpeer-review

    38 Citations (Scopus)


    The need to find new pattern recognition techniques that correctly classify complex structures has risen as an important field of research. A well-known solution to this problem, which has proven to be very powerful, is the use of mixture models. Mixture models are typically fitted using the expectation-maximization (EM) algorithm. Unfortunately, optimal results are not always achieved because the EM algorithm, iterative in nature, is only guaranteed to produce a local maximum. In this paper, a solution to this problem is proposed and tested in a complex structure where the classical EM algorithm normally fails. This, we will do by means of a genetic algorithm (GA) which will allow the system to combine different solutions in a stochastic search so as to produce better results. The reported results show the usefulness of this approach, and suggest how it can be successfully implemented. Two new algorithms are proposed. The first one is useful when a priori information of the observed data is not available. The second solution is useful for those cases where some knowledge of the structure of the data-set is known. This second solution has proven to converge faster than the first one, although the final results reached are very similar to each other.
    Original languageEnglish
    Pages (from-to)759-769
    JournalPattern Recognition Letters
    Issue number8
    Publication statusPublished - 1 Jan 2000


    • Computer vision
    • EM algorithm
    • Genetic algorithms
    • Mixture models


    Dive into the research topics of 'Learning mixture models using a genetic version of the EM algorithm'. Together they form a unique fingerprint.

    Cite this