• Journal of Internet Computing and Services
• /
• v.8 no.6
• /
• pp.43-53
• /
• 2007

Intrusion detection system(IDS) has recently evolved while combining signature-based detection approach with anomaly detection approach. Although signature-based IDS tools have been commonly used by utilizing machine learning algorithms, they only detect network intrusions with already known patterns, Ideal IDS tools should always keep the signature database of your detection system up-to-date. The system needs to generate the signatures to detect new possible attacks while monitoring and analyzing incoming network data. In this paper, we propose a new outlier cluster detection algorithm with density (or influence) function, Our method assumes that an outlier is a kind of cluster with similar instances instead of a single object in the context of network intrusion, Through extensive experiments using KDD 1999 Cup Intrusion Detection dataset. we show that the proposed method outperform the conventional outlier detection method using Euclidean distance function, specially when attacks occurs frequently.

• Journal of the Korea Institute of Information and Communication Engineering
• /
• v.17 no.4
• /
• pp.996-1004
• /
• 2013

In this paper, we propose two clustering schemes to prolong lifetime by improving unbalance of energy consumption among sensor nodes in wireless sensor networks. The first proposed scheme make up clusters according to density of sensor nodes in initial stage of wireless sensor networks for reducing energy consumption of wireless sensor networks. After the initial stage, a cluster header is selected by a relay scheme that determines a cluster header of next round among cluster members. by estimating of energy consumption of cluster members for improving unbalance of energy consumption among cluster members.

• Proceedings of the KIEE Conference
• /
• 2003.07d
• /
• pp.2546-2548
• /
• 2003

G-K 클러스터링이 가지는 우수한 클러스터 분류 성능에도 불구하고 데이터의 편중 및 분포 밀도에 의하여 클러스터링의 결과과 만족스럽지 못하는 경우가 발생한다. 제안된 방법에서는, G-K 클러스터링에서 데이터의 분포 및 밀도 등과 같은 다양한 조건에 대한 문제를 동시에 고려함으로써 클러스터링 결과를 개선한다. G-K 클러스터링에서 일부 파라미터의 수동적 파라미터 결정 방법을 Mountain 방법을 이용하여 능동적인 알고리즘으로 대치하여 클러스터 최적화 과정을 더욱 용이하게 한다. 이러한 클러스터링의 장점은 뉴로-퍼지 모델의 규칙 감소와 성능개선으로 나타나며 이를 시뮬레이션을 통하여 보이고자 한다.

• Journal of KIISE:Databases
• /
• v.35 no.6
• /
• pp.544-554
• /
• 2008

Histograms are popularly used for selectivity estimation in database systems. In conventional histogram methods, buckets return the approximated results based on the assumption that all objects in a bucket are uniformly distributed. However, the objects within the region of a query are not likely to be uniformly distributed. That is, there can be some skews (i.e., clusters) in the buckets, which may significantly degrade the accuracy of the histogram. The aim of this work is to enhance the accuracy of histograms. For this purpose, we propose a new two-dimensional histogram method considering clusters. The proposed method detects dense regions and exploits them for organizing buckets. Since the proposed method effectively reduces accuracy degradation caused by clusters, it can provide improved, robust accuracy against skewed data distributions. Through experiments, we show that the proposed method provides up to 74% improved performance compared with the conventional histogram.

• Journal of the Institute of Electronics Engineers of Korea CI
• /
• v.46 no.4
• /
• pp.79-86
• /
• 2009

In this paper, we propose a new clustering scheme that provides optimal data aggregation effect and reduces energy consumption of nodes by considering the density of nodes when forming clusters. Since the size of the cluster is determined to ensure optimal data aggregation rate, our scheme reduces transmission range and minimizes interference between clusters. Moreover, by clustering using locally adjacent nodes and aggregating data received from cluster members, we reduce energy consumption of nodes. Through simulation, we confirmed that energy consumption of the whole network is minimized and the sensor network life-time is extended. Moreover, we show that the proposed clustering scheme improves the performance of network compared to previous LEACH clustering scheme.

• Journal of IKEEE
• /
• v.23 no.2
• /
• pp.419-424
• /
• 2019

The FCM algorithm finds the optimal solution through iterative optimization technique. In particular, there is a difference in execution time depending on the initial center of clustering, the location of noise, the location and number of crowded densities. However, this method gradually updates the center point, and the center of the initial cluster is shifted to one side. In this paper, we propose a TI-FCM(Triangular Inequality-Fuzzy C-Means) clustering algorithm that determines the cluster center density by maximizing the distance between clusters using triangular inequality. The proposed method is an effective method to converge to real clusters compared to FCM even in large data sets. Experiments show that execution time is reduced compared to existing FCM.

• Proceedings of the Korea Information Processing Society Conference
• /
• 2001.10a
• /
• pp.755-758
• /
• 2001

본 논문에서는 서로 다른 크기의 클러스터에 대해서 효과적으로 데이터를 분류할 수 있는 내부클러스터를 이용한 개선된 FCM 알고리즘을 제안하였다. 내부클러스터는 평균내부거리 안쪽에 속하는 데이터 집합으로 클러스터의 크기와 밀도에 비례한다. 그러므로 이를 이용한 개선된 FCM 알고리즘은 기존의 FCM 알고리즘이 클러스터 크기가 다를 경우 퍼지분할과 중심탐색을 제대로 하지 못하는 문제점을 개선한 수 있다. 실험을 통하여 개선된 FCM 알고리즘이 분류 엔트로피에 의해 기존의 FCM 알고리즘 보다 더 좋은 결과를 나타냄을 알 수 있었다.

• Journal of the Korean Chemical Society
• /
• v.61 no.6
• /
• pp.328-338
• /
• 2017

The density functional theory(DFT) and ab initio calculations have been applied to investigate hydrogen interaction of $H_2O_3(H_2O)_n$ clusters(n=1-5). The structures, IR spectra, and H-bonding energies are calculated at various levels of theory. The $trans-H_2O_3$ monomer is predicted to be thermodynamically more stable than cis form at the CCSD(T)/cc-pVTZ level of theory. For clusters, the geometries are optimized at the MP2/cc-pVTZ level of theory. The binding energy of $H_2O_3-H_2O$ cluster is predicted to be -6.39 kcal/mol at the CCSD(T)//MP2/cc-pVTZ level of theory after zero-point vibrational energy (ZPVE) and basis set superposition error (BSSE) correction. This result implies that $H_2O_3$ is a stronger proton donor(acid) than either $H_2O$ or $H_2O_2$. The average binding energies per $H_2O$ are predicted to be 8.25 kcal/mol for n=2, 7.22 kcal/mol for n=3, 8.50 kcal/mol for n=4, and 8.16 kcal/mol for n=5.

• The Journal of the Acoustical Society of Korea
• /
• v.34 no.4
• /
• pp.303-309
• /
• 2015

Speaker diarization is the task of determining the speakers for unlabeled data, and DBSCAN (Density-Based Spatial Clustering of Applications with Noise) has been widely used in the field of speaker diarization for its simplicity and computational efficiency. One challenging issue, however, is that if different clusters in non-spatial dataset are adjacent to each other, over-clustering may occur which subsequently degrades the performance of DBSCAN. In this paper, we identify the drawbacks of DBSCAN and propose a new density clustering algorithm based on local distribution property around object. Variable density criterions for local density and spreadness of object are used for effective data clustering. We compare the proposed algorithm to DBSCAN in terms of clustering accuracy. Experimental results confirm that the proposed algorithm exhibits higher accuracy than DBSCAN without over-clustering and confirm that the new approach based on local density and object spreadness is efficient.

• Proceedings of the Korean Information Science Society Conference
• /
• 2008.06b
• /
• pp.240-245
• /
• 2008

컴퓨터 그래픽스에서 많은 광원들을 포함하는 장면을 사실적으로 렌더링하기 위해서는, 많은 양의 조명 계산을 수행해야 한다. 다수의 광원들로부터 빠르게 조명 계산을 하기 위해 많이 사용되는 기법 중에 몬테 카를로(Monte Carlo) 기법이 있다. 본 논문은 이러한 몬테 카를로(Monte Carlo) 기법을 기반으로, 다수의 광원들을 효과적으로 샘플링 할 수 있는 새로운 중요도 샘플링 기법을 제안한다. 제안된 기법의 두 가지 핵심 아이디어는 첫째, 장면 내에 다수의 광원이 존재하여도 어떤 특정 지역에 많은 영향을 주는 광원은 일부인 경우가 많다는 점이고 두 번째는 공간 일관성(spatial coherence)이 낮거나 그림자 경계 지역에 위치한 픽셀들은 영향을 받는 주요 광원이 서로 다르다는 점이다. 제안된 기법은 이러한 관찰에 착안하여 특정 지역에 광원이 기여하는 정도를 평가하고 이에 비례하게 확률 밀도 함수(PDF: Probability Density Function)를 결정하는 방법을 제안한다. 이를 위하여 이미지 공간상에서 픽셀들을 클러스터링(clustering)하고 클러스터 구조를 기반으로 대표 샘플을 선정한다. 선정된 대표 샘플들로부터 광원들의 기여도를 평가하고 이를 바탕으로 클러스터 단위의 확률 밀도 함수를 결정하여 최종 렌더링을 수행한다. 본 논문이 제안하는 샘플링 기법을 적용했을 때 전통적인 샘플링 방식과 비교하여 같은 샘플링 개수에서 노이즈(noise)가 적게 발생하는 좋은 화질을 얻을 수 있었다. 제안된 기법은 다수의 조명과 다양한 재질, 복잡한 가려짐이 존재하는 장면을 효과적으로 표현할 수 있다.