• 응용통계연구
• /
• 제32권1호
• /
• pp.149-160
• /
• 2019

본 연구는 2016년 7월부터 2017년 6월까지 인천 소재 A 대학교의 15분 단위의 일일 전기 사용량 시계열 데이터에 대해 functional data analysis 기법을 적용하여 군집화하고 각 군집의 특성을 파악하고 예측에 활용하고자 한다. 하루동안의 A 대학교의 전기 사용량은 패턴은 주중과 주말 에 큰 차이를 보이며 스플라인 기저함수로 FPCA 구한 후 이들에 대한 가우시안 분포의 혼합모형 기반 군집분석으로 3개의 군집화가 적절해 보인다. 각 군집에 대해 평균 함수, 확률밀도함수, 일들의 분포 등을 정리해 각 군집에 대한 정보와 특징을 보여준다.

• Communications for Statistical Applications and Methods
• /
• 제29권2호
• /
• pp.225-238
• /
• 2022

In this paper, we propose a clustering method for advanced metering infrastructure (AMI) data in Korea. As AMI data presents non-stationarity, we consider time-dependent frequency domain principal components analysis, which is a proper method for locally stationary time series data. We develop a new clustering method based on time-varying eigenvectors, and our method provides a meaningful result that is different from the clustering results obtained by employing conventional methods, such as K-means and K-centres functional clustering. Simulation study demonstrates the superiority of the proposed approach. We further apply the clustering results to the evaluation of the electricity price system in South Korea, and validate the reform of the progressive electricity tariff system.

• Structural Engineering and Mechanics
• /
• 제54권2호
• /
• pp.221-237
• /
• 2015

An integrated method is proposed for structural nonlinear damage detection based on time series analysis and the higher statistical moments of structural responses in this study. It combines the time series analysis, the higher statistical moments of AR model residual errors and the fuzzy c-means (FCM) clustering techniques. A few comprehensive damage indexes are developed in the arithmetic and geometric mean of the higher statistical moments, and are classified by using the FCM clustering method to achieve nonlinear damage detection. A series of the measured response data, downloaded from the web site of the Los Alamos National Laboratory (LANL) USA, from a three-storey building structure considering the environmental variety as well as different nonlinear damage cases, are analyzed and used to assess the performance of the new nonlinear damage detection method. The effectiveness and robustness of the new proposed method are finally analyzed and concluded.

• 대한원격탐사학회:학술대회논문집
• /
• 대한원격탐사학회 2008년도 International Symposium on Remote Sensing
• /
• pp.406-409
• /
• 2008

Clustering data streams has an importance over many applications like sensor networks. Existing hierarchical methods follow a semi fuzzy clustering that yields duplicate clusters. In order to solve the problems, we propose an extended online divisive agglomerative clustering on data streams. It builds a tree-like top-down hierarchy of clusters that evolves with data streams using geometric time frame for snapshots. It is an enhancement of the Online Divisive Agglomerative Clustering (ODAC) with a pruning strategy to avoid duplicate clusters. Our main features are providing update time and memory space which is independent of the number of examples on data streams. It can be utilized for clustering sensor data and network monitoring as well as web click streams.

• Communications for Statistical Applications and Methods
• /
• 제29권6호
• /
• pp.695-708
• /
• 2022

Recently, several studies have been conducted using state space model. In this study, a dynamic linear model with state space model form is applied to stock data. The monthly returns for 135 Korean stocks are fitted to a dynamic linear model, to obtain an estimate of the time-varying 𝛽-coefficient time-series. The model formula used for the return is a capital asset pricing model formula explained in economics. In particular, the transition equation of the state space model form is appropriately modified to satisfy the assumptions of the error term. k-shape clustering is performed to classify the 135 estimated 𝛽 time-series into several groups. As a result of the clustering, four clusters are obtained, each consisting of approximately 30 stocks. It is found that the distribution is different for each group, so that it is well grouped to have its own characteristics. In addition, a common pattern is observed for each group, which could be interpreted appropriately.

• 한국지능시스템학회논문지
• /
• 제19권5호
• /
• pp.595-602
• /
• 2009

시스템의 동작특성을 표현하는 퍼지 규칙들은 퍼지 클러스터링 기법에 매우 의존적이다. 만약, 클러스터링 기법의 분류 능력이 개선된다면, 그들에 의해 생성되는 퍼지 규칙과 식별되는 파라미터들이 보다 정밀해 질 수 있으므로 시스템의 성능이 개선될 수 있다. 따라서 본 논문에서는 분류능력이 강화된 새로운 계층 구조 클러스터링 알고리즘을 제안한다. 제안된 클러스터링 기법은 데이터 사이의 통계적 특성과 상관성을 고려하여 보다 정확하게 데이터들을 분류할 수 있도록 2개의 클러스터의 구조를 갖는다. 또한, 본 논문은 차분 데이터를 이용하여 원형 데이터의 패턴이나 규칙들이 명확하게 반영될 수 있도록 하며, 각각의 차분 데이터들의 다양한 특성을 고려할 수 있도록 다중 퍼지 시스템을 구현한다. 마지막으로, 제안된 기법들의 유효성을 다양한 비선형 시계열 데이터들의 예측을 통해 검증한다.

• 예술인문사회 융합 멀티미디어 논문지
• /
• 제7권8호
• /
• pp.305-314
• /
• 2017

Clustering is a method that collects data objects into groups based on their similary. Performance of the state-of-the-art clustering methods is different according to the data characteristics. There have been numerous studies that performed experiments to compare the accuracy of the state-of-the-art clustering methods by applying various kinds of datasets. A common problem of these studies is that they only consider clustering algorithms that yield the most accurate results for a particular dataset. They do not consider what factors affect the execution time of each clustering method and how they are affected. Nevertheless, execution time is an important factor in clustering performance if there is no significant difference in accuracy. In order to solve the problems of the existing research, through a series of experiments using various types of datasets, we compare the accuracy of four representative clustering methods. In addition, we perform practical clustering performance comparisons by deriving time complexity and identifying factors that influences to its performance.

• 전기학회논문지
• /
• 제60권7호
• /
• pp.1396-1403
• /
• 2011

It is not easy to analyze the strong nonlinear time series and effectively design a good prediction system especially due to the difficulties in handling the potential uncertainty included in data and prediction method. To solve this problem, a new design method for fuzzy prediction system is suggested in this paper. The proposed method contains the followings as major parts ; the first-order difference detection to extract the stable information from the nonlinear characteristics of time series, the fuzzy rule generation based on the hierarchically classifying clustering technique to reduce incorrectness of the system parameter identification, and the IT2TSK fuzzy logic system to reasonably handle the potential uncertainty of the series. In addition, the design of the multiple predictors is considered to reflect sufficiently the diverse characteristics concealed in the series. Finally, computer simulations are performed to verify the performance and the effectiveness of the proposed prediction system.

• Communications for Statistical Applications and Methods
• /
• 제17권3호
• /
• pp.377-390
• /
• 2010

In the case of time-series analysis, it is often more convenient to rely on the frequency domain than the time domain. Spectral density is the core of the frequency-domain analysis that describes autocorrelation structures in a time-series process. Possible ways to estimate spectral density are to compute a periodogram or to average the periodogram over some frequencies with (un)equal weights. This can be an attractive tool to measure the similarity between time-series processes. We employ the metrics based on a smoothed periodogram proposed by Park and Kim (2008) for the classification of different classes of time-series processes. We consider several lag windows with unequal weights instead of a modified Daniel's window used in Park and Kim (2008). We evaluate the performance under various simulation scenarios. Simulation results reveal that the metrics used in this study split the time series into the preassigned clusters better than do the raw-periodogram based ones proposed by Caiado et al. 2006. Our metrics are applied to an economic time-series dataset.

• 한국멀티미디어학회논문지
• /
• 제20권5호
• /
• pp.800-807
• /
• 2017

The Pattern Sequence-based Forecasting (PSF) is an approach to forecast the behavior of time series based on similar pattern sequences. The innovation of PSF method is to convert the load time series into a label sequence by clustering technique in order to lighten computational burden. However, it brings about a new problem in determining the number of clusters and it is subject to insufficient similar days occasionally. In this paper we proposed an adaption of the PSF method, which introduces a new clustering index to determine the number of clusters and imposes a threshold to solve the problem caused by insufficient similar days. Our experiments showed that the proposed method reduced the mean absolute percentage error (MAPE) about 15%, compared to the PSF method.