• 응용통계연구
• /
• 제18권3호
• /
• pp.661-671
• /
• 2005

본 연구는 반복분할(recursive partitioning)에 의한 군집화 방법을 개발하고 활용 예를 보인다. 노드 분리 기준으로는 Overall R-Square를 채택하였고 실용적인 노드 분리 결정 방법을 제안하였다. 이 방법은 연속형 자료에 대하여 나무 형태의 해석하기 쉬운 단순한 규칙을 제공하면서 동시에 변수선택기능을 제공한다. 환용 예로서 Fisher의 붓꽃데이터와 Telecom 사례에 적용해 보았다. K-평균 군집화와 다른 몇 가지 사항이 관측되었다.

• 응용통계연구
• /
• 제25권3호
• /
• pp.503-512
• /
• 2012

Detecting outliers in a linear regression model eventually fails when similar observations are classified differently in a sequential process. In such circumstances, identifying clusters and applying certain methods to the clustered data can prevent a failure to detect outliers and is computationally efficient due to the reduction of data. In this paper, we suggest to implement a clustering procedure for this purpose and provide examples that illustrate the suggested procedure applied to the Hadi-Simonoff (1993) method, reverse Hadi-Simonoff method, and Gentleman-Wilk (1975) method.

• 응용통계연구
• /
• 제23권2호
• /
• pp.367-374
• /
• 2010

As an alternative to the k-means clustering the k-spatial medians clustering has many good points because of advantages of spatial median. However, it has not been used a lot since it needs heavy computation. If the number of objects and the number of variables are large the computation time problem is getting serious. In this study we propose fast algorithm for the k-spatial medians clustering. Practical applicability of the algorithm is shown with some numerical studies.

• 산업경영시스템학회지
• /
• 제45권3호
• /
• pp.123-130
• /
• 2022

Pair trading is a statistical arbitrage investment strategy. Traditionally, cointegration has been utilized in the pair exploring step to discover a pair with a similar price movement. Recently, the clustering analysis has attracted many researchers' attention, replacing the cointegration method. This study tests a clustering-driven pair trading investment strategy in the Korean stock market. If a pair detected through clustering has a large spread during the spread exploring period, the pair is included in the portfolio for backtesting. The profitability of the clustering-driven pair trading strategies is investigated based on various profitability measures such as the distribution of returns, cumulative returns, profitability by period, and sensitivity analysis on different parameters. The backtesting results show that the pair trading investment strategy is valid in the Korean stock market. More interestingly, the clustering-driven portfolio investments show higher performance compared to benchmarks. Note that the hierarchical clustering shows the best portfolio performance.

• Communications for Statistical Applications and Methods
• /
• 제21권3호
• /
• pp.225-234
• /
• 2014

We propose a novel hierarchical clustering for distribution valued dissimilarities. Analysis of large and complex data has attracted significant interest. Symbolic Data Analysis (SDA) was proposed by Diday in 1980's, which provides a new framework for statistical analysis. In SDA, we analyze an object with internal variation, including an interval, a histogram and a distribution, called a symbolic object. In the study, we focus on a cluster analysis for distribution valued dissimilarities, one of the symbolic objects. A hierarchical clustering has two steps in general: find out step and update step. In the find out step, we find the nearest pair of clusters. We extend it for distribution valued dissimilarities, introducing a measure on their order relations. In the update step, dissimilarities between clusters are redefined by mixture of distributions with a mixing ratio. We show an actual example of the proposed method and a simulation study.

• Journal of the Korean Data and Information Science Society
• /
• 제18권4호
• /
• pp.927-935
• /
• 2007

Currently, Gyeongnam province is executing the social index survey every year to the provincials. But, this survey has the limit of the analysis as execution of the different survey per 3 year cycles. The solution of this problem is data fusion. Data fusion is the process of combining multiple data in order to provide information of tactical value to the user. But, data fusion doesn#t mean the ultimate result. Therefore, efficient analysis for the data fusion is also important. In this study, we present data fusion method of statistical survey data. Also, we suggest application methodology of association rule mining by clustering through data fusion of statistical survey data.

• Journal of Information Processing Systems
• /
• 제16권4호
• /
• pp.750-759
• /
• 2020

The wide application of various integrated location-based services (LBS social) and tourism application (app) has generated a large amount of trajectory space data. The trajectory data are used to identify popular tourist attractions with high density of tourists, and they are of great significance to smart service and emergency management of scenic spots. A hot spot analysis method is proposed, based on spatial clustering of trajectory stop points. The DBSCAN algorithm is studied with fast clustering speed, noise processing and clustering of arbitrary shapes in space. The shortage of parameters is manually selected, and an improved method is proposed to adaptively determine parameters based on statistical distribution characteristics of data. DBSCAN clustering analysis and contrast experiments are carried out for three different datasets of artificial synthetic two-dimensional dataset, four-dimensional Iris real dataset and scenic track retention point. The experiment results show that the method can automatically generate reasonable clustering division, and it is superior to traditional algorithms such as DBSCAN and k-means. Finally, based on the spatial clustering results of the trajectory stay points, the Getis-Ord Gi* hotspot analysis and mapping are conducted in ArcGIS software. The hot spots of different tourist attractions are classified according to the analysis results, and the distribution of popular scenic spots is determined with the actual heat of the scenic spots.

• 응용통계연구
• /
• 제8권1호
• /
• pp.39-50
• /
• 1995

기존의 집락분석은 집락화만을 목적으로 하기 때문에 분석이 끝나면 집락분석에 사용된 규칙을 보존하지 못하는 문제를 갖고 있다. 이러한 문제를 인간의 뇌의 성질을 연구하는 신경회로망 분야에서 사용하는 모형중 하나인 쿨롱 에너지 네트워크 모형을 변형 발전시켜서 해결하여 보았다. 이 모형을 이용한 분석의 실제 예를 보이고 기존의 기법들과의 비교를 통하여 거의 유사한 집락형성을 보여주고 있음을 보였다.

• 응용통계연구
• /
• 제33권4호
• /
• pp.511-526
• /
• 2020

단세포 RNA 시퀀싱 데이터(single-cell RNA-sequencing data, 이하 단세포 RNA 데이터)는 세포 조직으로부터 추출한 각 단세포 별 유전자의 신호를 기록한 데이터로, 세포 간의 이질성을 파악하는 것을 주요 목적으로 한다. 그러나 단세포 RNA 데이터는 샘플링 및 기술적인 한계로 인해 결측비율이 높고, 노이즈가 크다. 이러한 이유 때문에 기존의 군집화 방법을 적용하는 데에 한계가 존재한다. 본 논문에서는 단세포 RNA 데이터 분석에서 모티브를 얻어 스펙트럼 군집화(spectral clustering) 기반의 방법을 제안한다. 특히 유사도 행렬(similarity matrix) 계산에서 유전자 별로 가중치를 부여하여 기존의 단세포 데이터 분석 방법과 차별화하였다. 제안하는 군집화 방법은 유전자별 가중치를 부여함과 동시에 세포를 군집화한다. 군집화는 반복 알고리즘을 통해 제안하는 비볼록식(non-convex optimization)을 풀어 진행한다. 또한 실데이터 적용과 시뮬레이션을 통해 제안하는 군집화 방법이 기존의 방법보다 군집을 잘 구분하는 것을 보인다.

• Communications for Statistical Applications and Methods
• /
• 제6권1호
• /
• pp.53-68
• /
• 1999

To improve the slow convergence property of the steepest ascent type algorithm for continuous D-optimal design problems. we develop a new algorithm. We apply the nonlinear system of equations as the necessary condition of optimality and develop the two-point algorithm that solves the problem of clustering. Because of the nature of the steepest coordinate ascent algorithm avoiding the problem of clustering itself helps the improvement of convergence speed. The numerical examples show the performances of the new method is better than those of various steepest ascent algorithms.