In this paper, we firstly give a brief introduction of expectation maximization (EM) algorithm, and then discuss the initial value sensitivity of expectation maximization algorithm. Subsequently, we give a short proof of EM's convergence. Then, we implement experiments with the expectation maximization algorithm (We implement all the experiments on Gaussion mixture model (GMM) ). Our experiment...

II. Data Modeling Image segmentation is an essential step in many advanced imaging applications. Accurate segmentation is required for volume measurements, medical diagnosis, image guided procedures, and 3D rendering. In this paper, we have developed a new multiresolution algorithm that extends the well-known expectation maximization (EM) algorithm. The proposed algorithm is based on generating...

A sampling based EM algorithm is proposed. The algorithm tries to add some randomness to the likelihood function thereby increasing the chances of the convergence of the EM algorithm to the true maxima as against getting stuck in the local maximum of likelihood manifold. The experiments for the problem of parameter estimation for a mixture of two gaussians are presented. The improved accuracy i...

The EM algorithm heavily relies on the interpretation of observations as incomplete data but it does not have any control on the uncertainty of missing data. To effectively reduce the uncertainty of missing data, we present a regularized EM algorithm that penalizes the likelihood with the mutual information between the missing data and the incomplete data (or the conditional entropy of the miss...

The EM algorithm has been widely used in many learning or statistical tasks. However, since it requires multiple database scans, applying the EM algorithm to data streams is not straight forward. In this paper we propose an online EM algorithm which can deal with data streams. The algorithm utilizes a component reduction technique which reduces the number of components in a mixture model. A not...

Inference of network internal link characteristics has become an increasingly important issue for network monitor and network management. In this paper, an improved EM algorithm based on A* algorithm was proposed for network link delay distributions inference. We use A* algorithm to accelerate the convergence speed of EM algorithm. Experiment results show the improved EM algorithm is faster tha...

Average case performance of the deterministic annealing EM algorithm is evaluated for Gaussian mixture estimation problem under some additive noises. The data-averaged EM update equations with respect to hyperparameters are calculated analytically in the large data limit. We find that the EM algorithm strongly depends on the initial conditions. Moreover, by using our analysis, it becomes possib...

Segmentation of anatomical regions of brain is fundamental problem in medical image analysis. In general, segmentation is process in which object is divided into its constituents parts. In this paper, MRI of brain is divided into its constituent’s parts which may be used for medical diagnosis purpose. Consequently we used statistical approach to segment the brain image. Using this approach, we ...

the relation between single nucleotide polymorphisms (snps) and some diseases has been concerned by many researchers. also the missing snps are quite common in genetic association studies. hence, this article investigates the relation between existing snps in dnmt1 of human chromosome 19 with colorectal cancer. this article aims is to presents an imputation method for missing snps not at random...

The expectation maximization (EM) algorithm computes maximum likelihood estimates of unknown parameters in probabilistic models involving latent variables. More pragmatically speaking, the EM algorithm is an iterative method that alternates between computing a conditional expectation and solving a maximization problem, hence the name expectation maximization. We will in this work derive the EM ...