• 제어로봇시스템학회:학술대회논문집
• /
• 제어로봇시스템학회 2005년도 ICCAS
• /
• pp.1254-1259
• /
• 2005

In this paper, we proposed a learning algorithm for the neuro-fuzzy modeling using a learning rule to adapt clustering. The proposed algorithm includes the data partition, assigning the rule into the process of partition, and optimizing the parameters using predetermined threshold value in self-constructing algorithm. In order to improve the clustering, the learning method of neuro-fuzzy model is extended and the learning scheme has been modified such that the learning of overall model is extended based on the error-derivative learning. The effect of the proposed method is presented using simulation compare with previous ones.

• International Journal of Fuzzy Logic and Intelligent Systems
• /
• 제5권2호
• /
• pp.95-101
• /
• 2005

In this paper, we proposed a learning algorithm for the neuro-fuzzy modeling using a learning rule to adapt clustering. The proposed algorithm includes the data partition, assigning the rule into the process of partition, and optimizing the parameters using predetermined threshold value in self-constructing algorithm. In order to improve the clustering, the learning method of neuro-fuzzy model is extended and the learning scheme has been modified such that the learning of overall model is extended based on the error-derivative learning. The effect of the proposed method is presented using simulation compare with previous ones.

• 한국통신학회논문지
• /
• 제30권6C호
• /
• pp.490-496
• /
• 2005

본 논문에서는 퍼지 추론 시스템의 추론 정보를 이용하여 자율적으로 구조를 결정하는 클러스터링 기법을 제안한다. 제안된 기법은 주어진 입출력 데이터를 이용하여 자율적으로 클러스터의 수를 추정하고 동시에 이들 파라미터를 최적화한다. 일반적인 클러스터링 기법에서 볼 수 있었던 비교사학습을 교사학습으로 확장하여 클러스터 추정에 입출력 인과 관계를 고려한 학습을 실시하게 하여 전체 모델의 성능을 개선하고자 하였다. 출력 정보가 입력공간에서 클러스터링 학습에 적용됨으로써 클러스터링에서의 각 클래스의 구분 작업이 더 원활하게 이루어 질 수 있다. 모의실험을 통하여 기존의 연구 결과와 비교하여 제안된 기법의 유용성을 보인다.

• Journal of information and communication convergence engineering
• /
• 제8권6호
• /
• pp.609-617
• /
• 2010

This paper is concerned with the use of a selforganizing map (SOM) to estimate the desired channel states of an unknown digital communication channel for blind equalization. The modification of SOM is accomplished by using the Bayesian likelihood fitness function and the relation between the desired channel states and channel output states. At the end of each clustering epoch, a set of estimated clusters for an unknown channel is chosen as a set of pre-defined desired channel states, and used to extract the channel output states. Next, all of the possible desired channel states are constructed by considering the combinations of extracted channel output states, and a set of the desired states characterized by the maximal value of the Bayesian fitness is subsequently selected for the next SOM clustering epoch. This modification of SOM makes it possible to search the optimal desired channel states of an unknown channel. In simulations, binary signals are generated at random with Gaussian noise, and both linear and nonlinear channels are evaluated. The performance of the proposed method is compared with those of the "conventional" SOM and an existing hybrid genetic algorithm. Relatively high accuracy and fast search speed have been achieved by using the proposed method.

• 산업경영시스템학회지
• /
• 제38권3호
• /
• pp.56-63
• /
• 2015

Several fields of science have demanded large-scale workflow support, which requires thousands of CPU cores or more. In order to support such large-scale scientific workflows, high capacity parallel systems such as supercomputers are widely used. In order to increase the utilization of these systems, most schedulers use backfilling policy: Small jobs are moved ahead to fill in holes in the schedule when large jobs do not delay. Since an estimate of the runtime is necessary for backfilling, most parallel systems use user's estimated runtime. However, it is found to be extremely inaccurate because users overestimate their jobs. Therefore, in this paper, we propose a novel system for the runtime prediction based on workload-aware clustering with the goal of improving prediction performance. The proposed method for runtime prediction of parallel applications consists of three main phases. First, a feature selection based on factor analysis is performed to identify important input features. Then, it performs a clustering analysis of history data based on self-organizing map which is followed by hierarchical clustering for finding the clustering boundaries from the weight vectors. Finally, prediction models are constructed using support vector regression with the clustered workload data. Multiple prediction models for each clustered data pattern can reduce the error rate compared with a single model for the whole data pattern. In the experiments, we use workload logs on parallel systems (i.e., iPSC, LANL-CM5, SDSC-Par95, SDSC-Par96, and CTC-SP2) to evaluate the effectiveness of our approach. Comparing with other techniques, experimental results show that the proposed method improves the accuracy up to 69.08%.

• 전자공학회논문지C
• /
• 제35C권2호
• /
• pp.71-79
• /
• 1998

In this paper, we propose a new mART(modified ART) neural network by combining the winner neuron definition method of SOM(self-organizing map) and the real-time adaptive clustering function of ART(adaptive resonance theory) and construct it in a modular structure, for the purpose of organizing the feature maps of three dimensional targets. Being constructed in a modular structure, the proposed modular mART can effectively prevent the clusters from representing multiple classes and can be trained to organze two dimensional distortion invariant feature maps so as to recognize targets with three dimensional distortion. We also present the recognition result and self-organization perfdormance of the proposed modular mART neural network after carried out some experiments with 14 tank and fighter target models.

• 대한전기학회:학술대회논문집
• /
• 대한전기학회 2002년도 하계학술대회 논문집 B
• /
• pp.1172-1174
• /
• 2002

An adaptive fuzzy P + ID controller for variable speed operation of BLDC motor drives is presented in this paper. Generally, a conventional PID controller is most widely used in industry due to its simple control structure and ease of design. However, the PID controller suffers from the electrical machine parameter variations and disturbances. To improve the tracking performance for parameter and load variations, the controller proposed in this paper is constructed by using an adaptive fuzzy logic controller in place of the proportional term in a conventional PID controller. For implementing this controller, only one additional parameter has to be adjusted in comparison with the PID controller. An adaptive fuzzy controller applied to proportional term to achieve robustness against parameter variations has simple structure and computational simplicity. The controller based on optimal fuzzy logic controller has an self-tuning characteristics with clustering. Computer simulation results show the usefulness of the proposed controller.

• 정보과학회 논문지
• /
• 제44권9호
• /
• pp.966-975
• /
• 2017

사물인터넷 환경에서는 센서 노드가 사물로 의미가 확장되고 각 사물들은 자신만의 의사결정을 통해 사물간 정보 수집 및 공유가 가능하다. 따라서 노드의 데이터를 싱크노드 또는 중앙 서버로 전송하는 것을 목적으로 하는 WSN 정보 수집 방법을 사물인터넷 환경에 그대로 적용하는 것은 비효율적이다. 또한 기존 WSN 방법은 주변 사물들의 정보를 수집하는 과정에서 모든 사물이 정보 수집에 참여하는 방식으로 구성되어 있기 때문에 전송횟수 증가 등의 다양한 문제가 발생한다. 본 논문에서는 각 사물들 간의 에너지 효율적인 정보 공유를 위한 클러스터링 및 속성 기반 데이터 교환 방법을 제안한다. 제안하는 방법은 우선 각 사물이 생산할 수 있는 데이터의 속성을 부여하여 클러스터를 구성한다. 데이터 교환시에는 부여된 속성을 이용해 데이터를 생산할 수 있는 사물들과 통신하여 에너지 효율성을 향상시킨다. 성능평가를 위해 TOSSIM을 이용하여 네트워크 수명, 평균 에너지 소비량 등을 측정하였다.

• 천문학회보
• /
• 제44권2호
• /
• pp.63.3-63.3
• /
• 2019

Nearby young moving groups (NYMGs hereafter) are gravitationally unbound loose young stellar associations located within 100 pc of the Sun. Since NYMGs are crucial laboratories for studying low-mass stars and planets, intensive searches for NYMG members have been performed. For identification of NYMG members, various strategies and methods have been applied. As a result, the reliability of the members in terms of membership is not uniform, which means that a careful membership re-assessment is required. In this study, I developed a NYMG membership probability calculation tool based on Bayesian inference (Bayesian Assessment of Moving Groups: BAMG). For the development of the BAMG tool, I constructed ellipsoidal models for nine NYMGs via iterative and self-consistent processes. Using BAMG, memberships of claimed members in the literature (N~2000) were evaluated, and 35 per cent of members were confirmed as bona fide members of NYMGs. Based on the deficiency of low-mass members appeared in mass function using these bona fide members, low mass members from Gaia DR2 are identified. About 2000 new M dwarf and brown dwarf candidate members were identified. Memberships of ~70 members with RV from Gaia were confirmed, and the additional ~20 members were confirmed via spectroscopic observation. Not relying on previous knowledge about the existence of nine NYMGs, unsupervised machine learning analyses were applied to NYMG members. K-means and Agglomerative Clustering algorithms result in similar trends of grouping. As a result, six previously known groups (TWA, beta-Pic, Carina, Argus, AB Doradus, and Volans-Carina) were rediscovered. Three the other known groups are recognized as well; however, they are combined into two new separate groups (ThOr+Columba and TucHor+Columba).

• 조명전기설비학회논문지
• /
• 제21권7호
• /
• pp.75-83
• /
• 2007

일반적으로 변압기의 고장진단을 위해 IEC 코드법이 사용되지만, 이 방법은 가스비율이 규정된 범위 내에 존재하지 않거나 경계조건에 있는 경우 숙련된 진단 전문가에게 의뢰하지 않고는 정확한 고장의 원인을 판정하는데 어려움이 있다. 이러한 문제점을 해결하기 위하여 본 논문에서는 SOM을 이용한 전력용 변압기의 고장진단 알고리즘을 제안한다. 제안된 방법은 훈련 데이터의 경쟁학습을 통하여 자기 구성 맵을 구축한 후, 실증 데이터를 구축된 맵에 적용하여 고장의 진단이 이루어진다. 또한 클러스터링 기법에 의해 구축된 정상/고장모델과 정상 데이터를 비교함으로써 고장의 추이 및 열화정도를 분석한다. 제안된 방법의 유용성을 보이기 위한 실험결과에서 기존의 방법들에 비해 향상된 진단결과를 보임을 확인할 수 있었다.