• 응용통계연구
• /
• 제25권3호
• /
• pp.471-483
• /
• 2012

One of the most important problems in cluster analysis is the selection of variables that truly define cluster structure, while eliminating noisy variables that mask such structure. Brusco and Cradit (2001) present VS-KM(variable-selection heuristic for K-means clustering) procedure for selecting true variables for K-means clustering based on adjusted Rand index. This procedure starts with the fixed number of clusters in K-means and adds variables sequentially based on an adjusted Rand index. This paper presents an updated procedure combining the VS-KM with the automated K-means procedure provided by Kim (2009). This automated variable selection procedure for K-means clustering calculates the cluster number and initial cluster center whenever new variable is added and adds a variable based on adjusted Rand index. Simulation result indicates that the proposed procedure is very effective at selecting true variables and at eliminating noisy variables. Implemented program using R can be obtained on the website "http://faculty.knou.ac.kr/sskim/nvarkm.r and vnvarkm.r".

• 디지털산업정보학회논문지
• /
• 제9권1호
• /
• pp.87-93
• /
• 2013

MANET has the advantage of having the flexibility to build easily a network in a difficult situation that builds a wired network. But, data transmission errors by movement of nodes and eavesdropping by wireless communications have become a problem of security. Authentication service is the most essential in order to overcome these problems and operate network stably. In this paper, we propose 3-tiers authentication structure to exclude of malicious node and operate stable network through more systematic and thorough node authentication. After network is composed into a cluster, cluster head which play CA role is elected. Among these, the highest-CA is elected. The highest-CA receives certificates to cluster head and the cluster head evaluates trust value of their member nodes. Authentication technique which issues member node key is used. We compared PSS and experimented to evaluate performance of proposed scheme in this paper and efficiency of the proposed technique through experience was confirmed.

• Journal of information and communication convergence engineering
• /
• 제11권4호
• /
• pp.241-246
• /
• 2013

A classic document clustering technique may incorrectly classify documents into different clusters when documents that should belong to the same cluster do not have any shared terms. Recently, to overcome this problem, internal and external knowledge-based approaches have been used for text document clustering. However, the clustering results of these approaches are influenced by the inherent structure and the topical composition of the documents. Further, the organization of knowledge into an ontology is expensive. In this paper, we propose a new enhanced text document clustering method using non-negative matrix factorization (NMF) and WordNet. The semantic terms extracted as cluster labels by NMF can represent the inherent structure of a document cluster well. The proposed method can also improve the quality of document clustering that uses cluster labels and term weights based on term mutual information of WordNet. The experimental results demonstrate that the proposed method achieves better performance than the other text clustering methods.

• Bulletin of the Korean Chemical Society
• /
• 제24권6호
• /
• pp.695-702
• /
• 2003

Recent experimental and theoretical advances on the aromatic alcohol-water clusters are reviewed, focusing on the structure of the hydrogen bonding between the alcoholic OH group and the binding water molecules. The interplay of experimental observations and theoretical calculations for the elucidation of the structure is demonstrated for phenol-water, benzyl alcohol-water, substituted phenol-water, naphthol-water and tropolone -water clusters. Discussion is made on assigning the role (either proton-donating or -accepting) of the hydroxyl group by measuring the shifts of infrared frequency of the OH stretching mode in the cluster from that of bare aromatic alcohol for the experimental determination of the cluster structure.

• 한국통신학회논문지
• /
• 제37A권9호
• /
• pp.772-779
• /
• 2012

현재 MANET에서는 DOS나 DDOS 공격이 증가하고 있지만, MANET은 제한된 대역폭, 계산 자원 및 배터리 전원을 가진 노드들로 구성된 네트워크이기 때문에 기존의 역추적 메커니즘을 적용할 수가 없다. 따라서 MANET에서 역추적 기법을 적용 시킬 시에는 각 노드가 가지는 자원을 효율적으로 사용해야한다. 그러나 기존의 애드혹 네트워크에 적용한 역추적 기법은 클러스터링 영역에서 각 노드를 대표하는 Cluster head가 역추적 정보를 관리하기 때문에 Cluster head의 과부하로 노드의 수명을 단축시키는 문제점을 가지게 된다. 게다가, 다중홉 클러스터링일 경우, 하나의 Cluster head가 더 많은 노드를 관리하기 때문에 문제는 더욱 심각해진다. 본 논문에서는 이러한 Cluster head의 오버헤드를 줄이기 위한, 역추적 정보를 관리하기 위하여 트리구조의 깊이를 이용한 TNA(Traceback against Network Attacks)를 제안한다.

• Journal of Microbiology and Biotechnology
• /
• 제15권5호
• /
• pp.1120-1124
• /
• 2005

Pseudomonas putida KCTC 1639 was newly identified as a potential producer of biodegradable medium chain length polyhydroxyalkanoates. It exhibited a carbon assimilation pattern quite different from other known P. putida strains, but a more similar pattern with P. oleovorans, which assimilates the carbon sources mainly through ${\beta}$-oxidation rather than the fatty acid biosynthesis pathway. The PHA synthesis gene cluster from P. putida KCTC1639 was composed of two gene loci; the PHA synthase gene locus and granule-associated gene locus, which were cloned and deposited in the GenBank under accession numbers AY286491 and AY750858 as a new nucleotide sequence, respectively. The molecular structure and amino acid homology of the new gene cluster were compared with those from Pseudomonas species, including other P. putida strains and P. oleovorans, and a higher than $90\%$ homology was observed.

• 한국환경복원기술학회지
• /
• 제26권4호
• /
• pp.1-16
• /
• 2023

The purpose of this study is to understand the vegetation structure of the Corylopsis coreana habitats in South Korea and their correlation with environmental factors, in order to provide basic data for preparing conservation. A total of 40 vegetation survey plots were established in around Southern region, Gangwon-do and Gyeonggi-do Province. The cluster analysis revealed four distinct clusters: Quercus mongolica-Pinus densiflora (QmPd), Castanea crenata (Cc), Quercus mongolica (Qm), and Pinus koraiensis (Pk). Among them, the QmPd cluster was selected as the representative community. Soil texture analysis that most areas consisted of loamy soil and were distributed on the northern aspects. The plantation, represented by Pk, exhibited a high content of exchangeable aluminum (5.227±0.342 mg/kg), suggesting the need for soil improvement and monitoring in these habitats. The canopy openness (forest gap) ranged from 11% to 21%, indicating a relatively closed canopy in many survey plots. Non-metric multidimensional scaling analysis indicated heterogeneous species composition between the QmPd cluster and the Pk cluster (total R² = 0.608).

• 한국육종학회지
• /
• 제42권1호
• /
• pp.11-22
• /
• 2010

The population structure of a domesticated species is influenced by the natural history of the populations of its pre-domesticated ancestors, as well as by the breeding system and complexity of breeding practices implemented by humans. In the genetic and population structure analysis of 122 South Asia collections using 29 simple sequence repeat (SSR) markers, 362 alleles were detected, with an average of 12.5 per locus. The average expected heterozygosity and polymorphism information content (PIC) for each SSR locus were 0.74 and 0.72,respectively. The model-based structure analysis revealed the presence of three clusters with the 91.8% (shared > 75%) membership, with 8.2% showing admixture. The genetic distances of Clusters 1-3 were 0.55, 0.56, and 0.68, respectively. Polymorphic information content followed the same trend (Cluster 3 had the highest value and Cluster 1 had smallest value), with genetic distances for each cluster of 0.52, 0.52, and 0.65, respectively. This result could be used for supporting rice breeding programs in South Asia countries.

• 전자공학회논문지CI
• /
• 제41권3호
• /
• pp.37-45
• /
• 2004

본 논문에서는 이동 에드-혹 네트워크에서 조합가중치 클러스터링 알고리즘을 적용하여 강건하고 신뢰성 있는 클러스터 기반의 그룹 멀티캐스트 방식을 제안한다. 에드-혹 네트워크는 고정된 통신 하부 구조의 도움 없이 이동 단말기로만 구성된 무선 네트워크이다. 제한된 대역폭과 높은 이동성으로 인하여 에드-혹 네트워크에서의 라우팅 프로토콜은 강건하고, 간단하면서 에너지 소비를 최소화하여야 한다. WCGM(Weighted Cluster Group Multicast)방식은 조합 가중치 다중 클러스터 기반 구조를 이용하고 기존의 FGMP(Forwarding Group Multicast Protocol)방식의 장점인 제한적인 플러딩에 의한 데이터 전달방식은 유지하면서 클러스터 헤드 선출 시 조합가중치를 적용한다. 이것은 안정적이며 강건한 데이터 전달 구조를 가지기 때문에 데이터 전달 구조를 유지하기 위한 오버헤드(Overhead)와 데이터 전달을 위한 오버헤드를 모두 줄이는 효과를 시뮬레이션을 통하여 검증하였다.

• 대한전자공학회논문지TC
• /
• 제49권8호
• /
• pp.84-97
• /
• 2012

무선 센서 네트워크에서 클러스터 구조는 부하균형, 에너지 절약, 분산 키 관리와 같은 중요한 이점들을 유발한다. 네트워크를 클러스터 구조로 변경하기 위해서 센서들은 배치된 후에 클러스터 구성 프로토콜을 실행해야 한다. 이때 일부 노드들이 공격자에 의해 오염될 수 있으며 이들이 클러스터 구성 프로토콜에 따르지 않으면 임의의 클러스터에서 멤버십의 비일치가 발생한다. 이러한 경우에 클러스터들은 분할되게 되며 이에 따라 클러스터 수는 증가하고 클러스터내의 멤버수가 줄어드는 문제점이 발생한다. 본 논문에서는 오염노드들의 프로토콜 비협조에 잘 대처하는 기법을 제안한다. 먼저, 제안방법은 클러스터 내에서 임의의 두 노드 간에 거리가 최대 2홉이 되는 2홉 클러스터들을 생성한다. 게다가 제안방법은 오염노드에 의한 클러스터들의 분할을 막기 위해 2홉 거리의 노드에 의한 검증을 이용한다. 또한 제안방법은 노드들의 에너지 소모를 감소시키기 위하여 주로 방송전송들을 이용한다. 실험결과들은 제안방법이 다른 방법에 비해 클러스터 수를 감소시키며 좀 더 안전하고 에너지 효율적임을 보여주었다.