• Journal of the Institute of Electronics Engineers of Korea SC
• /
• v.46 no.6
• /
• pp.1-8
• /
• 2009

Detecting entry and exit zones in a view covered by multiple cameras is an essential step to determine the topology of the camera setup, which is critical for achieving and sustaining the accuracy and efficiency of multi-camera surveillance system. In this paper, a graph theoretic clustering method is proposed to detect zones using data points which correspond to entry and exit events of objects in the camera view. The minimum spanning tree (MST) is constructed by associating the data points. Then a set of well-formed clusters is sought by removing inconsistent edges of the MST, based on the concepts of the cluster balance and the cluster density defined in the paper. Experimental results suggest that the proposed method is effective, even for sparsely elongated clusters which could be problematic for expectation-maximization (EM). In addition, comparing to the EM-based approaches, the number of data required to obtain stable outcome is relatively small, hence shorter learning period.

• The KIPS Transactions:PartC
• /
• v.17C no.2
• /
• pp.181-189
• /
• 2010

HAP(High Altitude Platform) is a stationary aerial platform positioned in the stratosphere between 17Km and 22Km height and it could act as an MBS (Mobile Base Station). HAP based Network has advantages of both satellite system and terrestrial communication system. In this paper we study the deploy of multiple HAP MBS that can provides efficient communication for users. For this study, EM(Expectation Maximization) clustering algorithm is used to cluster terrestrial mobile nodes. The object of this paper is improving EM algorithm into the clustering algorithm for efficiency in variety aspects considering distance between mobile terminal units and speed of mobile terminal units, and estimating performance of HAP MBS deploy technique with use of improved EM algorithm using RWP (Random Waypoint) node mobility.

• Journal of the Korean Institute of Intelligent Systems
• /
• v.26 no.6
• /
• pp.493-497
• /
• 2016

Infrared (IR) imaging is capable of detecting targets that are not visible at night, thus it has been widely used for the security and defense system. However, the quality of the IR image is often degraded by low resolution and noise corruption. This paper addresses target segmentation with the IR image. Multiple regions of interest (ROI) are extracted by the multi-level segmentation and targets are segmented from the individual ROI. Each level of the multi-level segmentation is composed of a k-means clustering algorithm an expectation-maximization (EM) algorithm, and a decision process. The k-means clustering algorithm initializes the parameters of the Gaussian mixture model (GMM) and the EM algorithm iteratively estimates those parameters. Each pixel is assigned to one of clusters during the decision. This paper proposes the selection and the merging of the extracted ROIs. ROI regions are selectively merged in order to include the overlapped ROI windows. In the experiments, the proposed method is tested on an IR image capturing two pedestrians at night. The performance is compared with conventional methods showing that the proposed method outperforms others.

• Proceedings of the Korea Water Resources Association Conference
• /
• 2017.05a
• /
• pp.45-45
• /
• 2017

상이한 자연현상으로 발생된 자료들은 때때로 통계적으로 다른 특성을 가지는 경우가 있다. 이런 자료들은 다른 두 개 이상의 모집단에서 자료가 발생한 것으로 가정할 수 가 있다. 기존에 널리 사용되어온 분포형 모형의 경우 단일한 모집단으로부터 자료가 발생한다는 가정하에서 개발된 모형들로 위에서 언급한 자료들을 적절히 모의할 수 없다. 이런 상이한 모집단에서 발생된 자료를 모형화 하기 위해서 혼합분포모형(mixture distribution)이 개발되었다. 홍수나 가뭄 등과 같은 극치 사상의 경우 다양한 자연현상들로부터 발생하기에 혼합분포모형을 적용할 경우 보다 정확한 모의가 가능하다. 혼합분포모형은 두 개 이상의 비혼합분포모형들을 가중합하여 만들어진다. 혼합 분포모형의 형태로 인하여 기존의 분포형 모형의 매개변수 추정 모형으로 널리 사용되던 최우도법 (maximum likelihood method), 모멘트법(method of moment), 확률가중모멘트법 (probability weighted moment method) 등을 이용하여 혼합분포모형의 매개변수를 추정하는 것이 용이 하지 않다. 혼합분포모형의 매개변수 추정 방법으로는 Expectation-Maximization (EM) 알고리즘, Meta-Heuristic Maximum Likelihood (MHML) 방법, Markov Chain Monte Carlo (MCMC) 방법 등이 적용되고 있다. 현재까지 수자원 분야에서 사용되는 극치 자료를 혼합분포모형을 이용하여 모의할 때 매개변수 추정방법에 따른 특성에 대한 연구가 진행되지 않았다. 본 연구에서는 우리나라 연최대강우량 자료를 이용하여 혼합분포모형의 매개변수 추정방법 (EM 알고리즘, MHML 방법, MCMC 방법) 들의 특성들을 비교 분석하였다. 혼합분포모형으로는 Gumbel-Gumbel 혼합분포 모형을 적용하였다. 본 연구의 결과는 향후 혼합분포모형을 이용한 연구에 좋은 기초자료로 사용될 수 있을 것으로 판단된다.

• Journal of the Korea Society of Computer and Information
• /
• v.27 no.12
• /
• pp.161-169
• /
• 2022

With the development of mobile devices and global positioning systems, various location-based services can be utilized, which collects user's location information and provides services based on it. In this process, there is a risk of personal sensitive information being exposed to the outside, and thus Geo-indistinguishability (Geo-Ind), which protect location privacy of LBS users by perturbing their true location, is widely used. However, owing to the data perturbation mechanism of Geo-Ind, it is hard to accurately obtain the density distribution of LBS users from the collection of perturbed location data. Thus, in this paper, we aim to develop a novel method which enables to effectively compute the user density distribution from perturbed location dataset collected under Geo-Ind. In particular, the proposed method leverages Expectation-Maximization(EM) algorithm to precisely estimate the density disribution of LBS users from perturbed location dataset. Experimental results on real world datasets show that our proposed method achieves significantly better performance than a baseline approach.

• The Journal of the Acoustical Society of Korea
• /
• v.26 no.8
• /
• pp.390-396
• /
• 2007

In this letter, we propose a novel approach to improve the performance of speech/music classification for the selectable mode vocoder(SMV) of 3GPP2 using the Gaussian mixture model(GMM) which is based on the expectation-maximization(EM) algorithm. We first present an effective analysis of the features and the classification method adopted in the conventional SMV. And then feature vectors which are applied to the GMM are selected from relevant Parameters of the SMV for the efficient speech/music classification. The performance of the proposed algorithm is evaluated under various conditions and yields better results compared with the conventional scheme of the SMV.

• Journal of the Korean Operations Research and Management Science Society
• /
• v.23 no.4
• /
• pp.111-130
• /
• 1998

We propose an imperfect repair model which depends on external effects quantified by covariates. The model is based on the Brown-Proschan imperfect repair model wherefrom the probability of perfect repair is represented by a function of covariates. We are motivated by deficiency of the BP model whose stationarity prevents us from predicting dynamically the time to next failure according to external condition. Five types of function for the probability of perfect repair are proposed. This article also presents a procedure for estimating the parameter of the function for the probability of perfect repair, as well as the inherent lifetime distribution of the system, based on consecutive inter-failure times and the covariates. The estimation procedure is based on the expectation-maximization principle which is suitable to incomplete data problems. focusing on the maximization step, we derive some theorems which guarantee the existence of the solution. A Monte Carlo study is also performed to illustrate the prediction power of the model as well as to show reasonable properties of the estimates. The model reduces significantly the mean square error of the in-sample prediction. so it can be utilized in real fields for evaluating and maintaining repairable systems.

• Journal of the Korean Data and Information Science Society
• /
• v.27 no.3
• /
• pp.587-598
• /
• 2016

In predicting an outcome of election using a variety of methods ahead of the election, non-response is one of the major issues. Therefore, to address the non-response issue, a variety of methods of non-response imputation may be employed, but the result of forecasting tend to vary according to methods. In this study, in order to improve electoral forecasts, we studied a model based method of non-response imputation attempting to apply the Monte Carlo Expectation Maximization (MCEM) algorithm, introduced by Wei and Tanner (1990). The MCEM algorithm using maximum likelihood estimates (MLEs) is applied to solve the boundary solution problem under the non-ignorable non-response mechanism. We performed the simulation studies to compare estimation performance among MCEM, maximum likelihood estimation, and Bayesian estimation method. The results of simulation studies showed that MCEM method can be a reasonable candidate for non-response model estimation. We also applied MCEM method to the Korean presidential election exit poll data of 2012 and investigated prediction performance using modified within precinct error (MWPE) criterion (Bautista et al., 2007).

• Journal of Korean Institute of Industrial Engineers
• /
• v.33 no.1
• /
• pp.52-60
• /
• 2007

Estimating parameters of the lifetime distribution is investigated when field failure data are not completelyreported. To take into account the reality and the accuracy of the estimates in such a case, the failure reportingprobability is incorporated in estimating parameters, Firstly, method of maximum likelihood estimate (MLE) isused to estimate parameters of the lifetime distribution when failure reporting probability is known, Secondly,Expectation and Maximization (EM) algorithm is used to estimate the failure reporting probability and parame-ters of the lifetime distribution simultaneously when failure reporting probability is unknown. For both cases,procedures of estimation are illustrated for single Weibull distribution and mixed Weibull distribution. Simula-tion results show that MLE obtained by the proposed method is more accurate than the conventional MLE.

• Korea Journal of Hospital Management
• /
• v.10 no.4
• /
• pp.75-97
• /
• 2005

The research objective is to empirically investigate the compensation packages maximizing the utilities of internal customers by applying the market segmentation theory. Data was collected from four Korean hospitals in Seoul, Busan and Gyunggi-do. The research is designed to seek the compensation package maximizing the utility of doctors by mixture regression model, which has been applied as latent structure and other type of finite mixture models from various academic fields since early 1980s. The mixture regression model shows the optimal segments number and fuzzy classification for each observation by EM(expectation-maximization algorism). The finite mixture regression model is to unmix the sample, to identify the groups, and to estimate the parameters of the density function underlying the observed data within each group. The doctors were segmented into 5 groups by their preference for the compensation package. The results of this study imply that the utility of doctors increases with differentiated compensation package segmented by their preference.