• The Journal of Korean Institute of Communications and Information Sciences
• /
• v.32 no.9C
• /
• pp.861-866
• /
• 2007

A clustering algorithm for Gaussian Probability Distribution Function (GPDF) data called Centroid Neural Network with a Bhattacharyya Kernel (BK-CNN) is proposed in this paper. The proposed BK-CNN is based on the unsupervised competitive Centroid Neural Network (CNN) and employs a kernel method for data projection. The kernel method adopted in the proposed BK-CNN is used to project data from the low dimensional input feature space into higher dimensional feature space so as the nonlinear problems associated with input space can be solved linearly in the feature space. In order to cluster the GPDF data, the Bhattacharyya kernel is used to measure the distance between two probability distributions for data projection. With the incorporation of the kernel method, the proposed BK-CNN is capable of dealing with nonlinear separation boundaries and can successfully allocate more code vector in the region that GPDF data are densely distributed. When applied to GPDF data in an image classification probleml, the experiment results show that the proposed BK-CNN algorithm gives 1.7%-4.3% improvements in average classification accuracy over other conventional algorithm such as k-means, Self-Organizing Map (SOM) and CNN algorithms with a Bhattacharyya distance, classed as Bk-Means, B-SOM, B-CNN algorithms.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2019.01a
• /
• pp.29-30
• /
• 2019

본 논문에서는 다중 컨트롤러가 존재하는 분산 SDN 환경에서 과도한 제어 메시지로 인한 과부하된 컨트롤러의 부하를 줄이기 위하여 이주할 스위치를 K-means 군집화와 Harmony Search(HS)를 기반으로 선정 하는 기법을 제안하였다. 기존에 HS를 이용하여 이주할 스위치를 선택하는 기법이 제시되었으나, 시간 소모에 비하여 정확도가 부족한 단점이 있다. 또한 Harmony Memory(HM) 구축을 위해 메모리 소모 또한 크다. 이를 해결하기 위하여 본 논문에서는 유클리드 거리를 기반으로 하는 K-means 군집화를 이용하여 이주할 스위치를 골라내어 HM의 크기를 줄이고 이주 효율을 향상 시킨다.

• Journal of Navigation and Port Research
• /
• v.48 no.3
• /
• pp.164-170
• /
• 2024

The light buoy, a floating structure at sea, is prone to drifting due to external factors such as oceanic weather. This makes it imperative to monitor for any loss or displacement of buoys. In order to address this issue, the Ministry of Oceans and Fisheries aims to issue alerts for buoy displacement by analyzing historical buoy position data to detect patterns. However, periodic lifting inspections, which are conducted every two years, disrupt the buoy's location pattern. As a result, new patterns need to be analyzed after each inspection for location monitoring. In this study, buoy position data from various periods were analyzed using convex hull and distance-based clustering algorithms. In addition, the optimal data collection period was identified in order to accurately recognize buoy location patterns. The findings suggest that a nine-week data collection period established stable location patterns, explaining approximately 89.8% of the variance in location data. These results can improve the management of light buoys based on location patterns and aid in the effective monitoring and early detection of buoy displacement.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2018.05a
• /
• pp.481-483
• /
• 2018

AR은 VR과 달리 실세계 공간의 객체에 대한 서비스를 제공하므로 서비스 개발을 방해하는 많은 요인들이 발생한다. 이를 보완하기위해 비주얼 마커, SLAM, 객체인식 등 여러 AR 기술이 존재한다. 본 논문은 AR 기술 중에서 객체인식의 정확도 향상을 위해 지역가변 이진화(Local variable binarization)와 색상의 군집화를 사용해서 이미지에서 객체를 추출하는 방법을 제안한다. 지역 가변화는 픽셀을 순차적으로 읽어 들이면서 픽셀 주위의 값의 평균을 구하고, 이 값을 해당 픽셀의 임계 값으로 사용하는 알고리즘이다. 픽셀마다 주위 색상 값에 의해 임계 값이 변화되므로 윤곽선 표현이 기존의 이진화보다 뚜렷이 나타난다. 색상의 군집화는 객체의 중요색상과 배경의 중요색상을 중심으로 유사한 색상끼리 군집화 하는 것이다. 객체 내에서 가장 많이 나온 값과 객체 외에 가장 많이 나온 값을 각 각 기준으로 색조와 채도의 값을 Euclidean 거리를 사용해 객체의 색상과 배경 색상을 분리했다.

• Proceedings of the Korean Society of Computer Information Conference
• /
• 2015.01a
• /
• pp.321-324
• /
• 2015

본 논문에서는 SNS에 존재하는 특정 제품과 브랜드 또는 기업에 대한 평가, 의견, 느낌, 사용 후기 등의 소비자 생각을 수집하여 기업에서 향후 신제품 개발이나 시장 진출 및 확대 등의 경영활동에 활용할 수 있도록 SNS 빅데이터를 문석하고, 이를 활용하여 보다 소집단화 되고 개인화 되어가는 Micro-Trend 중심의 마케팅 활동을 할 수 있는 Micro-Targeting 관련 분석 정보를 제공 모델링하는 것을 제안한다. 본 연구에서는 SNS 데이터의 수집, 저장, 분석에 대한 내용을 다루고 있으며, 특히 마이크로타겟팅을 위한 정보를 머하웃(Mahout)의 유클리드 거리 기반의 유사도와 K-평균 군집 알고리즘을 활용하여 구현하고자 하였다.

• Journal of KIISE:Software and Applications
• /
• v.35 no.9
• /
• pp.547-553
• /
• 2008

We propose a two-stage document layout segmentation method. At the first stage, as top-down segmentation, morphological distance map algorithm extracts a collection of rectangular regions from a given input image. This preliminary result from the first stage is employed as input parameters for the process of next stage. At the second stage, a machine-learning algorithm is adopted RBF network, one of neural networks based on statistical model, is selected. In order for constructing the hidden layer of RBF network, a data clustering technique bared on the self-organizing property of Kohonen network is utilized. We present a result showing that the supervised neural network, trained by 300 number of sample data, improves the preliminary results of the first stage.

• Smart Media Journal
• /
• v.9 no.3
• /
• pp.122-129
• /
• 2020

The possibility that an asymptotic coronavirus-19 infected person around the world is not aware of his infection and can spread it to people around him is still a very important issue in that the public is not free from anxiety and fear over the spread of the epidemic. In this paper, the K-means clustering algorithm and deep learning-based crowd aggregation were proposed to determine the possibility of contact with confirmed cases of infectious diseases. As a result of 300 iterations of all input learning images, the PSNR value was 21.51, and the final MAE value for the entire data set was 67.984. This means the average absolute error between observations and the average absolute error of fewer than 4,000 people in each CCTV scene, including the calculation of the distance and infection rate from the confirmed patient and the surrounding persons, the net group of potential patient movements, and the prediction of the infection rate.

• Journal of Internet Computing and Services
• /
• v.12 no.2
• /
• pp.103-111
• /
• 2011

Clustering methods is divided into hierarchical clustering, partitioning clustering, and more. If the amount of documents is huge, it takes too much time to cluster them in hierarchical clustering. In this paper we deal with K-Means algorithm that is one of partitioning clustering and is adequate to cluster so many documents rapidly and easily. We propose the new method of selecting initial seeds in K-Means algorithm. In this method, the initial seeds have been selected that are positioned as far away from each other as possible.

• Journal of the Korean Society for Aeronautical & Space Sciences
• /
• v.50 no.8
• /
• pp.523-530
• /
• 2022

Since the Multi-UAV system for various missions is more complex than a single UAV, an efficient formation control method is required. In wide-area search mission, there is a need for a distributed control for flexible formation that has a low burden of communication and computation and enables autonomous formation between UAVs. This paper proposes a flexible formation operation method that considers the swarm formation, the bank alignment formation, and the formation movement to expand the scan area and improve search performance. The algorithm has a vibration characteristic of the second-order system for a relative distance and can design an algorithm through parameter tuning. In addition, we converted control commands to suit conventional UAV systems and demonstrated the performance of algorithms for a formation and movement of a formation through simulation.

• The Journal of the Korea Contents Association
• /
• v.6 no.12
• /
• pp.113-120
• /
• 2006

This paper proposes DBSCAN-SI that is an algorithm for clustering with influences of spatial objects. DBSCAN-SI that is extended from existing DBSCAN and DBSCAN-W converts from non-spatial properties to the influences of spatial objects during the spatial clustering. It increases probability of inclusion to the cluster according to the higher the influences that is affected by the properties used in clustering and executes the clustering not only respect the spatial distances, but also volume of influences. For the perspective of specific property-centered, the clustering technique proposed in this paper can makeup the disadvantage of existing algorithms that exclude the objects in spite of high influences from cluster by means of being scarcely close objects around the cluster.