An online (recursive) algorithm is proposed that estimates the parameters of the mixture and that simultaneously selects the number of components to search for the maximum a posteriori (MAP) solution and to discard the irrelevant components.
Abstract:
There are two open problems when finite mixture densities are used to model multivariate data: the selection of the number of components and the initialization. In this paper, we propose an online (recursive) algorithm that estimates the parameters of the mixture and that simultaneously selects the number of components. The new algorithm starts with a large number of randomly initialized components. A prior is used as a bias for maximally structured models. A stochastic approximation recursive learning algorithm is proposed to search for the maximum a posteriori (MAP) solution and to discard the irrelevant components.
TL;DR: An efficient adaptive algorithm using Gaussian mixture probability density is developed using Recursive equations to constantly update the parameters and but also to simultaneously select the appropriate number of components for each pixel.
TL;DR: This work presents recursive equations that are used to constantly update the parameters of a Gaussian mixture model and to simultaneously select the appropriate number of components for each pixel and presents a simple non-parametric adaptive density estimation method.
TL;DR: A heuristic for searching for the optimal component to insert in the greedy learning of gaussian mixtures is proposed and can be particularly useful when the optimal number of mixture components is unknown.
TL;DR: This paper provides an overview of the different representative clustering methods and several clustering validations indices and approaches to automatically determine the number of clusters.
TL;DR: In this article, the authors proposed an online kernel density estimation (KDE) method, which maintains and updates a non-parametric model of the observed data, from which the KDE can be calculated.
TL;DR: In this article, a new estimate minimum information theoretical criterion estimate (MAICE) is introduced for the purpose of statistical identification, which is free from the ambiguities inherent in the application of conventional hypothesis testing procedure.
TL;DR: In this paper, the problem of selecting one of a number of models of different dimensions is treated by finding its Bayes solution, and evaluating the leading terms of its asymptotic expansion.
TL;DR: In this paper, the problem of selecting one of a number of models of different dimensions is treated by finding its Bayes solution, and evaluating the leading terms of its asymptotic expansion.
Q1. What have the authors contributed in "Recursive unsupervised learning of finite mixture models - pattern analysis and machine intelligence, ieee transactions on" ?
In this paper, the authors propose an online ( recursive ) algorithm that estimates the parameters of the mixture and that simultaneously selects the number of components.
Q2. How many trials did the algorithm perform?
With random initialization, the authors performed 100 trials and the new algorithm was always able to find the correct solution while simultaneously estimating the parameters of the mixture and selecting the number of components.
Q3. How many iterations is required to fit the mixture?
The batch algorithm from [6] is fitting the mixture and selecting 11, 12, or 13 components using typically 300 to 400 iterations for a 900 samples data set.
Q4. What is the main limitation of the EM algorithm?
one of the serious limitations of the EM algorithm is that it can end up in a poor local maximum if not properly initialized.
Q5. What is the heuristic used to estimate the parameters of a mixture?
In [3] and [6], an efficient heuristic was used to simultaneously estimate the parameters of a mixture and select the appropriate number of its components.
Q6. How many components are used in the batch algorithm?
The batch algorithms assume a known number of components: three for the “Three Gaussians” and the “Iris” data, 13 for the “Shrinking Spiral,” and four for the “Enzyme” data set.
Q7. What is the optimum log-likelihood for the available data?
Most of the practical model selectiontechniques are based on maximizing the following type of criteria:JðM;~ ðMÞÞ ¼ log pðX ;~ ðMÞÞ P ðMÞ: ð3ÞHere, log pðX ;~ ðMÞÞ is the log-likelihood for the available data.