• Communications for Statistical Applications and Methods
• /
• v.15 no.1
• /
• pp.1-11
• /
• 2008

For a large data set the high computational cost of estimating the parameters of normal mixtures with the conventional EM algorithm is crucially impedimental in applying the algorithm to the areas requiring high speed computation such as real-time speech recognition. Simulations show that the binned EM algorithm, being compared to the standard one, significantly reduces the cost of computation without loss in accuracy of the final estimates.

• Journal of Korean Society for Quality Management
• /
• v.30 no.4
• /
• pp.68-77
• /
• 2002

EM algorithm has good convergence rate for numerical procedures which converges on very small step. In the case of proportion estimation in a mixed distribution which has very big incomplete data or of update of new data continuously, however, EM algorithm highly depends on a initial value with slow convergence ratio. There have been many studies to improve the convergence rate of EM algorithm in estimating the proportion parameter of a mixed data. Among them, dynamic EM algorithm by Hurray Jorgensen and Titterington algorithm by D. M. Titterington are proven to have better convergence rate than the standard EM algorithm, when a new data is continuously updated. In this paper we suggest dynamic EM algorithm and Titterington algorithm for the estimation of a mixed Poisson distribution and compare them in terms of convergence rate by using a simulation method.

• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.34 no.4C
• /
• pp.397-402
• /
• 2009

Accurate estimation of time-selective fading channel is a difficult problem in OFDM(Orthogonal Frequency Division Multiplexing) system. There are many channel estimation algorithms that are very weak in noisy channel. For solving this problem, we use EM (Expectation-Maximization) algorithm for iterative optimization of the data. We propose an EM-LPC algorithm to estimate the time-selective fading. The proposed algorithm improves of the BER performance compared to EM based channel estimation algorithm and reduces the iteration number of the EM loop. We simulated the uncoded system. If coded system use the EM-LPC algorithm, the performance are enhanced because of the coding gain. The EM-LPC algorithm is able to apply to another communication system, not only OFDM systems. The image processing of the medical instruments that the demand of accurate estimation can also use the proposed algorithm.

• The KIPS Transactions:PartD
• /
• v.13D no.3 s.106
• /
• pp.369-376
• /
• 2006

This paper presents a new method of significantly improving conventional Bayesian statistical text classifier by incorporating accelerated EM(Expectation Maximization) algorithm. EM algorithm experiences a slow convergence and performance degrade in its iterative process, especially when real online-textual documents do not follow EM's assumptions. In this study, we propose a new accelerated EM algorithm with uncertainty-based selective sampling, which is simple yet has a fast convergence speed and allow to estimate a more accurate classification model on Naive Bayesian text classifier. Experiments using the popular Reuters-21578 document collection showed that the proposed algorithm effectively improves classification accuracy.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.25 no.6
• /
• pp.523-528
• /
• 2015

In this paper, we propose a new multiresolutional and dynamic approach of the EM algorithm. EM is a very popular and powerful clustering algorithm. EM, however, has problems that indexes multiresolution data and requires a priori information on a proper number of clusters in many applications, To solve such problems, the proposed EM algorithm can impose a multiresolution kd-tree structure in the E-step and allocates a cluster based on sequential data. To validate clusters, we use a merge criteria for cluster merging. We demonstrate the proposed EM algorithm outperforms for texture image segmentation.

• Communications for Statistical Applications and Methods
• /
• v.29 no.2
• /
• pp.277-286
• /
• 2022

Due to the uniquely slow convergence speed of the EM algorithm, it suffers form a lot of processing time until the desired deconvolution image is obtained when the image is large. To cope with the problem, in this paper, an immediate solution of the EM algorithm is provided under the Gaussian image model. It is derived by finding the recurrent formular of the EM algorithm and then substituting the results repeatedly. In this paper, two types of immediate soultion of image deconboution by EM algorithm are provided, and both methods have been shown to work well. It is expected that it free the processing time of image deconvolution because it no longer requires an iterative process. Based on this, we can find the statistical properties of the restored image at specific iterates. We demonstrate the effectiveness of the proposed method through a simple experiment, and discuss future concerns.

• Communications for Statistical Applications and Methods
• /
• v.19 no.1
• /
• pp.157-168
• /
• 2012

The EM algorithm is the most important tool to obtain the maximum likelihood estimator in finite mixture models due to its stability and simplicity. However, its convergence rate is often slow because the conventional EM algorithm is based on a large missing data space. Several techniques have been proposed in the literature to reduce the missing data space. In this paper, we review existing methods and propose a new EM algorithm for Gaussian mixtures, which reduces the missing data space while preserving the stability of the conventional EM algorithm. The performance of the proposed method is evaluated with other existing methods via simulation studies.

• Proceedings of the KIEE Conference
• /
• 2003.07d
• /
• pp.2615-2617
• /
• 2003

The Kalman filter is a recursive Linear Estimator for the linear dynamic systems(LDS) affected by two different noises called process noise and measurement noise both of which are uncorrelated white. The Expectation Maximization(EM) algorithm is employed in this paper as a preprocessor to reinforce the effectiveness of Kalman estimator. Particularly, we focus on the relation between Kalman filter and EM algorithm in the LDS. In this paper, we propose a new algorithm to improve the performance on the parameter estimation via EM algorithm, which improves the overall process of Kalman filtering. Since Kalman filter algorithm not only needs the system parameters but also is very sensitive the initial state conditions, the initial conditions decided through EM turns out to be very effective. In experiments, the computer simulation results ate provided to demonstrate the superiority of the proposed algorithm.

• Proceedings of the Korean Information Science Society Conference
• /
• 2001.04b
• /
• pp.529-531
• /
• 2001

We present an expectation-maximization algorithm for principal component analysis via orthogonalization. The algorithm finds actual principal components, whereas previously proposed EM algorithms can only find principal subspace. New algorithm is simple and more efficient thant probabilistic PCA specially in noiseless cases. Conventional PCA needs computation of inverse of the covariance matrices, which makes the algorithm prohibitively expensive when the dimensions of data space is large. This EM algorithm is very powerful for high dimensional data when only a few principal components are needed.

• International Journal of Internet, Broadcasting and Communication
• /
• v.6 no.1
• /
• pp.16-22
• /
• 2014

Clustering methods such as k-means and EM are the group of classification and pattern recognition, which are used in management science and literature search widely. In this paper, k-means and EM algorithm are compared the performance using by Weka. The winning Lottery numbers of 567 cases are experimented for our study and presentation. Processing speed of the k-means algorithm is superior to the EM algorithm, which is about 0.08 seconds faster than the other. As the result it is summerized that EM algorithm is better than K-means algorithm with comparison of accuracy, precision and recall. While K-means is known to be sensitive to the distribution of data, EM algorithm is probability sensitive for clustering.