Genetic-based EM algorithm for learning Gaussian mixture models

Abstract

We propose a genetic-based expectation-maximization (GA-EM) algorithm for learning Gaussian mixture models from multivariate data. This algorithm is capable of selecting the number of components of the model using the minimum description length (MDL) criterion. Our approach benefits from the properties of Genetic algorithms (GA) and the EM algorithm by combination of both into a single procedure. The population-based stochastic search of the GA explores the search space more thoroughly than the EM method. Therefore, our algorithm enables escaping from local optimal solutions since the algorithm becomes less sensitive to its initialization. The GA-EM algorithm is elitist which maintains the monotonic convergence property of the EM algorithm. The experiments on simulated and real data show that the GA-EM outperforms the EM method since: 1) We have obtained a better MDL score while using exactly the same termination condition for both algorithms. 2) Our approach identifies the number of components which were used to generate the underlying data more often than the EM algorithm.

Automatic Classification of Embryonic Fruit Fly Gene Expression Patterns

“…The preprocessing (i) consists of a shading correction and a contrast optimization method resulting in a "'clean image"'. The shape segmentation (ii) is executed on the clean image by computing the magnitude of the gradient followed by a fragmentation of the gradient feature space with the Gaussian Mixture Model [15,16] instead of using a static threshold. The result is a probability map describing the probability for the membership to one of the classes (background, embryo).…”

A new random approach for initialization of the multiple restart EM algorithm for Gaussian model-based clustering

“…Another random initialization method [10,30] uses mixing proportions equal to 1=K, selects random feature vectors as component means, and employs a spherical covariance matrix equal to 0:1r 2 I, where I is the identity matrix and…”

“…In yet another approach to the GMM parameter estimation problem, the EM algorithm is combined with some global optimization method, e.g.. an evolutionary algorithm [1,30] or a particle swarm optimizer [7]. However, global optimizers have high computational demands and this approach is limited to moderately sized datasets.…”

“…Each digit is represented by a 16-dimensional feature vector. In [30] a subset of this dataset describing the first five digits was used for GMM learning. To evaluate the learning algorithm in a more challenging setting, in the presented experiments the complete dataset (K ¼ 10) consisting of 10992 feature vectors was used.…”

Semi Supervised Image Segmentation Using Optimal Color Seed Selection

“…The genetic algorithm [16,17] then creates a population of solutions and applies genetic operators such as mutation and crossover to evolve the solutions in order to find the best one(s).…”