• Journal of the Korean Statistical Society
• /
• 제33권3호
• /
• pp.255-269
• /
• 2004

This paper develops log-density estimation for directional data. The methodology is to use expansions with respect to spherical harmonics followed by estimating the unknown parameters by maximum likelihood. Minimax rates of convergence in terms of the Kullback-Leibler information divergence are obtained.

• Journal of the Korean Society for Industrial and Applied Mathematics
• /
• 제7권1호
• /
• pp.63-77
• /
• 2003

Some inequalities for the $Csisz{\acute{a}}r\;{\Phi}-divergence$ and applications for the Kullback-Leibler, $R{\acute{e}}nyi$, Hellinger and Bhattacharyya distances in Information Theory are given.

• 응용통계연구
• /
• 제24권6호
• /
• pp.1271-1284
• /
• 2011

본 논문에서는 위치와 척도모수가 모두 알려지지 않은 역가우스분포에 대한 적합도 검정으로 기존에 개발된 엔트로피 기반 검정을 확장한 쿨백-라이블러 정보 기반 적합도 검정을 소개한다. 역가우스분포에 대한 단순 또는 복합 영가설을 검정하기 위한 4가지 형태의 검정통계량을 제시하고 검정통계량의 계산에 사용할 표본크기에 따른 윈도크기와 기각값을 모의실험을 통해 결정하여 표의 형태로 제공한다. 검정력 분석을 위해 수행한 모의실험의 결과에서 위치와 척도모수가 모두 알려진 역가우스분포에 대한 쿨백-라이블러 정보 기반 적합도 검정은 모든 대립분포와 표본크기에서 EDF 검정들보다 좋은 검정력을 가지는 것으로 나타난다. 위치모수 또는 척도모수만 알려진 역가우스분포에 대한 쿨백-라이블러 정보 기반 적합도 검정은 모든 대립분포에 대해서 표본크기가 커짐에 따라 검정력이 증가하는 경향을 보인다. 위치와 척도모수가 모두 알려지지 않은 역가우스분포에 대한 쿨백-라이블러 정보 기반 적합도 검정은 대체적으로 엔트로피 기반 검정과 비슷한 수준의 검정력을 보이는 것으로 나타나고 이 결과를 통해서 두 검정은 동일함을 확인할 수 있다.

• The Journal of the Acoustical Society of Korea
• /
• 제27권2E호
• /
• pp.57-63
• /
• 2008

Stethoscopic auscultation is still one of the primary tools for the diagnosis of heart diseases due to its easy accessibility and relatively low cost. It is, however, a difficult skill to acquire. Many research efforts have been done on the automatic classification of heart sound signals to support clinicians in heart sound diagnosis. Recently, hidden Markov models (HMMs) have been used quite successfully in the automatic classification of the heart sound signal. However, in the classification using HMMs, there are so many heart sound signal types that it is not reasonable to assign a new class to each of them. In this paper, rather than constructing an HMM for each signal type, we propose to build an HMM for a set of acoustically-similar signal types. To define the classes, we use the KL (Kullback-Leibler) distance between different signal types to determine if they should belong to the same class. From the classification experiments on the heart sound data consisting of 25 different types of signals, the proposed method proved to be quite efficient in determining the optimal set of classes. Also we found that the class determination approach produced better results than the heuristic class assignment method.

• Journal of the Korean Statistical Society
• /
• 제29권1호
• /
• pp.1-8
• /
• 2000

A spectrum minimizing the frequency-domain Kullback-Leibler information number has been proposed and used to modify a spectrum estimate. Some numerical examples have illustrated the KL spectrum estimate is superior to the initial estimate, i.e., the autocovariances obtained by the inverse Fourier transformation of the KL spectrum estimate are closer to the sample autocovariances of the given observations than those of the initial spectrum estimate. Also, it has been shown that a Gaussian autoregressive process associated with the KL spectrum is the closest in the timedomain Kullback-Leibler sense to a Gaussian white noise process subject to given autocovariance constraints. In this paper a corresponding conditional probability theorem is presented, which gives another rationale to the KL spectrum.

• 한국정보과학회논문지:데이타베이스
• /
• 제37권6호
• /
• pp.285-291
• /
• 2010

본 연구에서는 나이브 베이시안 학습의 환경에서 속성의 가중치를 계산하는 새로운 방식을 제안한다. 기존 방법들이 속성에 가중치를 부여하는 방식인데 반하여 본 연구에서는 한걸음 더 나아가 속성의 값에 가중치를 부여하는 새로운 방식을 연구하였다. 이러한 속성값의 가중치를 계산하기 위하여 Kullback-Leibler 함수를 이용하여 가중치를 계산하는 방식을 제안하였고 이러한 가중치들의 특성을 분석하였다. 제안된 알고리즘은 다수의 데이터를 이용하여 속성 가중치 방식과 비교하였고 대부분의 경우에 더 좋은 성능을 제공함을 알 수 있었다.

• 한국통계학회:학술대회논문집
• /
• 한국통계학회 2003년도 추계 학술발표회 논문집
• /
• pp.233-238
• /
• 2003

Goodness of fit test statistics based on the information discrepancy have been shown to perform very well (Vasicek 1976, Dudewicz and van der Meulen 1981, Chandra et al 1982, Gohkale 1983, Arizona and Ohta 1989, Ebrahimi et al 1992, etc). Although the test is well defined for the non-censored case, censored case has not been discussed in the literature. Therefore we consider a goodness of fit test based on the partial Kullback-Leibler(KL) information with the type II censored data. We derive the partial KL information of the null distribution function and a nonparametric distribution function, and establish a goodness of fit test statistic. We consider the exponential and normal distributions and made Monte Calro simulations to compare the test statistics with some existing tests.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제5권6호
• /
• pp.1133-1146
• /
• 2011

Cognitive radio (CR) is considered one of the most promising next-generation communication systems; it has the ability to sense and make use of vacant channels that are unused by licensed users. Reliable detection of the licensed users' signals is an essential element for a CR network. Cooperative spectrum sensing (CSS) is able to offer better sensing performance as compared to individual sensing. The presence of malicious users who falsify sensing data can severely degrade the sensing performance of the CSS scheme. In this paper, we investigate a secure CSS scheme, based on the Kullback-Leibler Divergence (KL-divergence) theory, in order to identify malicious users and mitigate their harmful effect on the sensing performance of CSS in a CR network. The simulation results prove the effectiveness of the proposed scheme.

• Communications for Statistical Applications and Methods
• /
• 제21권2호
• /
• pp.125-134
• /
• 2014

Kullback-Leibler (KL) information is a measure of discrepancy between two probability density functions. However, several nonparametric density function estimators have been considered in estimating KL information because KL information is not well-defined on the empirical distribution function. In this paper, we consider the KL information of the equilibrium distribution function, which is well defined on the empirical distribution function (EDF), and propose an EDF-based goodness of fit test statistic. We evaluate the performance of the proposed test statistic for an exponential distribution with Monte Carlo simulation. We also extend the discussion to the censored case.

• Journal of the Korean Statistical Society
• /
• 제24권2호
• /
• pp.453-470
• /
• 1995

We describe a constrained optimal classification rule for the case when the prior probability of an observation belonging to one of the two populations is unknown. This is done by suggesting a balanced design for the classification experiment and constructing the optimal rule under the balanced design condition. The rule si characterized by a constrained minimization of total risk of misclassification; the constraint of the rule is constructed by the process of equation between Kullback-Leibler's directed divergence measures obtained from the two population conditional densities. The efficacy of the suggested rule is examined through two-group normal classification. This indicates that, in case little is known about the relative population sizes, dramatic gains in accuracy of classification result can be achieved.