• The Transactions of the Korea Information Processing Society
• /
• v.7 no.7
• /
• pp.2246-2253
• /
• 2000

In this paper, we show that one can enhance the performance of blind segmentation of phoneme boundaries by adopting the knowledge of Korean syllabic structure and the regions of voiced/unvoiced sounds. eh proposed method consists of three processes : the process to extract candidate phoneme boundaries, the process to detect boundaries of voiced/unvoiced sounds, and the process to select final phoneme boundaries. The candidate phoneme boudaries are extracted by clustering method based on similarity between two adjacent clusters. The employed similarity measure in this a process is the ratio of the probability density of adjacent clusters. To detect he boundaries of voiced/unvoiced sounds, we first compute the power density spectrum of speech signal in 0∼400 Hz frequency band. Then the points where this paper density spectrum variation is greater than the threshold are chosen as the boundaries of voiced/unvoiced sounds. The final phoneme boundaries consist of all the candidate phoneme boundaries in voiced region and limited number of candidate phoneme boundaries in unvoiced region. The experimental result showed about 40% decrease of insertion rate compared to the blind segmentation method we adopted.

• Journal of Korean Institute of Industrial Engineers
• /
• v.40 no.3
• /
• pp.257-266
• /
• 2014

The semiconductor manufacturing industry is managed by a number of parameters from the FAB which is the initial step of production to package test which is the final step of production. Various methods for prediction for the quality and yield are required to reduce the production costs caused by a complicated manufacturing process. In order to increase the accuracy of quality prediction, we have to extract the significant features from the large amount of data. In this study, we propose the method for extracting feature from the cell level data of probe test process using OPTICS which is one of the density-based clustering to improve the prediction accuracy of the quality of the assembled chips that will be placed in a package test. Two features extracted by using OPTICS are used as input variables of quality prediction model because of having position information of the cell defect. The package test progress for chips classified to the correct quality grade by performing the improved prediction method is expected to bring the effect of reducing production costs.

• Journal of KIISE:Computing Practices and Letters
• /
• v.16 no.3
• /
• pp.346-350
• /
• 2010

Since sensor nodes have restriction of using resources in Zigbee network, number of study on improving efficiency is currently ongoing[1]. Clustering mechanism based on hierarchy structure provides a prevention of duplicated information and a facility of a network expansion[2]. however overheads can occurs when the cluster header is elected and the election of a incorrect cluster header causes to use resources inefficiently. In this paper, we propose that the cluster header election mechanism using distances between nodes and density of nodes in accordance with the operation of the central processing system in which the sync nodes are having information of location and energy with respect to general nodes based on hierachy clustering mechanism.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2021.10a
• /
• pp.186-188
• /
• 2021

본 논문에서는 초음파 영상에서 DBSCAN(Density-based spatial clustering of applications with noise)과 FCM 클러스터링 기반 양자화 기법을 적용하여 결절종을 추출하는 방법을 제안한다. 본 논문에서는 초음파 영상 촬영 시 좌우 상단의 지방층 영역과 하단 영역의 명암도가 어두운 영역을 잡음 영역으로 설정한다. 그리고 초음파 영상에 퍼지스트레칭 기법을 적용하여 잡음 영역을 최대한 제거 한 후에 ROI 영역을 추출한다. 추출된 ROI 영역에서 밀도 분포를 분석하기 위하여 히스토그램을 분석한 후에 DBSCAN을 적용하여 초음파 영상에서 결절종 후보에 해당되는 명암도를 추출한다. 추출한 후보 명암도를 대상으로 FCM 클러스터링 기법을 적용한다. FCM을 적용하는 단계에서 결절종의 저에코 혹은 무에코의 특징을 이용하여 클러스터 중심 값이 가장 낮은 클러스터를 양자화 한 후에 라벨링 기법을 적용시켜 결절종의 후보 객체를 추출한다. 제안된 결절종 추출 방법의 성능을 분석하기 위해 전문의가 결절종 영역을 표기한 초음파 영상과 표기되지 않은 초음파 영상 120쌍을 대상으로 DBSCAN, FCM, 그리고 제안된 방법 간의 성능을 비교 분석하였다. 제안된 방법에서는 120개의 초음파 영상에서 106개 결절종 영역이 추출되었고 FCM 기법에서는 80개가 추출되었고 DBSCAN에서는 36개가 추출되었다. 따라서 제안된 방법이 결절종 추출에 효율적인 것을 확인하였다.

• Journal of the Korea Society of Computer and Information
• /
• v.11 no.2 s.40
• /
• pp.43-52
• /
• 2006

LRecently, with increasing interest in ubiquitous computing, knowledge discovery method is needed with consideration of the efficiency and the effectiveness of wide range and various forms of data. Spatial Characterization which extends former characterization method with consideration of spatial and non-spatial property enables to find various form of knowledge in spatial region. The previous spatial characterization methods have the problems as follows. Firstly, former study shows the problem that the result of searched knowledge is unable to perform the multiple spatial analysis. Secondly, it is unable to secure the useful knowledge search since it searches the limited spatial region which is allocated by the user. Thus, this study suggests spatial characterization which applies to density based clustering.

• The Journal of The Korea Institute of Intelligent Transport Systems
• /
• v.19 no.5
• /
• pp.82-96
• /
• 2020

The Demand Responsive Transit (DRT) system is the flexible public transport service that determines the route and schedule of the service vehicles according to users' requests. With increasing importance of public transport systems in urban areas, the development of stable and fast routing algorithms for DRT has become the goal of many researches over the past decades. In this study, a new heuristic method is proposed to generate fast and efficient routes for multiple vehicles using demand clustering and destination demand priority searching method considering the imbalance of users' origin and destination demands. The proposed algorithm is tested in various demand distribution scenarios including random, concentration and directed cases. The result shows that the proposed method reduce the drop of service ratio due to an increase in demand density and save computation time compared to other algorithms. In addition, compared to other clustering-based algorithms, the walking cost of the passengers is significantly reduced, but the detour time and in-vehicle travel time of the passenger is increased due to the detour burden.

• Proceedings of the Korean Institute of Information and Commucation Sciences Conference
• /
• 2019.05a
• /
• pp.341-343
• /
• 2019

As the generation of high concentration particulate matter increases, much attention is focused on the prediction of particulate matter. Particulate matter refers to particulate matter less than $10{\mu}m$ diameter in the atmosphere and is affected by weather changes such as temperature, relative humidity and wind speed. Therefore, various studies have been conducted to analyze the correlation with weather information for particulate matter prediction. However, the nonlinear time series distribution of particulate matter increases the complexity of the prediction model and can lead to inaccurate predictions. In this paper, we try to mitigate the nonlinear characteristics of particulate matter by using cluster algorithm and classification algorithm of machine learning. The machine learning algorithms used are agglomerative clustering, density-based spatial clustering of applications with noise(DBSCAN).

• Journal of the Korean Society of Surveying, Geodesy, Photogrammetry and Cartography
• /
• v.32 no.3
• /
• pp.245-251
• /
• 2014

As delineating market areas of retail businesses has become an interesting topic in marketing field, Lee and Lee recently suggested a noteworthy method, which applied the hydrological analysis of geographical information system (GIS), based on Christaller's central place theory. They used a digital elevation model (DEM) which inverted the kernel density of retail businesses, which was measured by using bandwidths of pre-determined 500, 1000 and 5000 m, respectively. In fact, their method is not a fully data-based approach in that they used pre-determined kernel bandwidths, however, this paper has been planned to improve Lee and Lee's method by using a kind of data-based approach of the L-index that describes clustering level of point feature distribution. The case study is implemented to automobile-related retail businesses in Seoul, Korea with selected Kernel bandwidths, 1211.5, 2120.2 and 7067.2 m from L-index analysis. Subsequently, the kernel density is measured, the density DEM is created by inverting it, and boundaries of market areas are extracted. Following the study, analysis results are summarized as follows. Firstly, the L-index can be a useful tool to complement the Lee and Lee's market area analysis method. At next, the kernel bandwidths, pre-determined by Lee and Lee, cannot be uniformly applied to all kinds of retail businesses. Lastly, the L-index method can be useful for analyzing the space structure of market areas of retail businesses, based on Christaller's central place theory.

• The KIPS Transactions:PartD
• /
• v.15D no.4
• /
• pp.485-494
• /
• 2008

Huge amounts of various web items such as keywords, images, and web pages are being made widely available on the Web. The popularities of such web items continuously change over time, and mining temporal patterns in popularities of web items is an important problem that is useful for several web applications. For example, the temporal patterns in popularities of search keywords help web search enterprises predict future popular keywords, enabling them to make price decisions when marketing search keywords to advertisers. However, presence of millions of web items makes it difficult to scale up previous techniques for this problem. This paper proposes an efficient method for mining temporal patterns in popularities of web items. We treat the popularities of web items as time-series, and propose gapmeasure to quantify the similarity between the popularities of two web items. To reduce the computation overhead for this measure, an efficient method using the Fast Fourier Transform (FFT) is presented. We assume that the popularities of web items are not necessarily following any probabilistic distribution or periodic. For finding clusters of web items with similar popularity trends, we propose to use a density-based clustering algorithm based on the gap measure. Our experiments using the popularity trends of search keywords obtained from the Google Trends web site illustrate the scalability and usefulness of the proposed approach in real-world applications.

• Journal of the Korea Convergence Society
• /
• v.10 no.11
• /
• pp.159-164
• /
• 2019

Recent developments in ICT technology have increased interest in the care and health of pets such as dogs and cats. In this paper, cluster analysis was performed based on the component data of pet food to be used in various fields of the pet industry. For cluster analysis, the similarity was analyzed by analyzing the correlation between components of 300 dogs and cats in the market. In this paper, clustering techniques such as Hierarchical, K-Means, Partitioning around medoids (PAM), Density-based, Mean-Shift are clustered and analyzed. We also propose a personalized recommendation system for pets. The results of this paper can be used for personalized services such as feed recommendation system for pets.