• 한국지능정보시스템학회:학술대회논문집
• /
• 한국지능정보시스템학회 2001년도 The Pacific Aisan Confrence On Intelligent Systems 2001
• /
• pp.30-33
• /
• 2001

The number of hidden neurons of the feed-forward neural networks is generally decided on the basis of experience. The method usually results in the lack or redundancy of hidden neurons, and causes the shortage of capacity for storing information of learning overmuch. This research proposes a new method for optimizing the number of hidden neurons bases on information entropy, Firstly, an initial neural network with enough hidden neurons should be trained by a set of training samples. Second, the activation values of hidden neurons should be calculated by inputting the training samples that can be identified correctly by the trained neural network. Third, all kinds of partitions should be tried and its information gain should be calculated, and then a decision-tree correctly dividing the whole sample space can be constructed. Finally, the important and related hidden neurons that are included in the tree can be found by searching the whole tree, and other redundant hidden neurons can be deleted. Thus, the number of hidden neurons can be decided. In the case of building a neural network with the best number of hidden units for tea quality evaluation, the proposed method is applied. And the result shows that the method is effective

• Smart Structures and Systems
• /
• 제10권4_5호
• /
• pp.443-458
• /
• 2012

The structural health monitoring (SHM) benchmark study on optimal sensor placement problem for the instrumented Canton Tower has been launched. It follows the success of the modal identification and model updating for the Canton Tower in the previous benchmark study, and focuses on the optimal placement of vibration sensors (accelerometers) in the interest of bettering the SHM system. In this paper, the sensor placement problem for the Canton Tower and the benchmark model for this study are first detailed. Then an information entropy based sensor placement method with the purpose of damage detection is proposed and applied to the benchmark problem. The procedure that will be implemented for structural damage detection using the data obtained from the optimal sensor placement strategy is introduced and the information on structural damage is specified. The information entropy based method is applied to measure the uncertainties throughout the damage detection process with the use of the obtained data. Accordingly, a multi-objective optimal problem in terms of sensor placement is formulated. The optimal solution is determined as the one that provides equally most informative data for all objectives, and thus the data obtained is most informative for structural damage detection. To validate the effectiveness of the optimally determined sensor placement, damage detection is performed on different damage scenarios of the benchmark model using the noise-free and noise-corrupted measured information, respectively. The results show that in comparison with the existing in-service sensor deployment on the structure, the optimally determined one is capable of further enhancing the capability of damage detection.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제16권8호
• /
• pp.2587-2605
• /
• 2022

In order to efficiently detect community structure in complex networks, community detection algorithms can be designed from the perspective of node similarity. However, the appropriate parameters should be chosen to achieve community division, furthermore, these existing algorithms based on the similarity of common neighbors have low discrimination between node pairs. To solve the above problems, a noval community detection algorithm using closeness similarity based on common neighbor node clustering entropy is proposed, shorted as CSCDA. Firstly, to improve detection accuracy, common neighbors and clustering coefficient are combined in the form of entropy, then a new closeness similarity measure is proposed. Through the designed similarity measure, the closeness similar node set of each node can be further accurately identified. Secondly, to reduce the randomness of the community detection result, based on the closeness similar node set, the node leadership is used to determine the most closeness similar first-order neighbor node for merging to create the initial communities. Thirdly, for the difficult problem of parameter selection in existing algorithms, the merging of two levels is used to iteratively detect the final communities with the idea of modularity optimization. Finally, experiments show that the normalized mutual information values are increased by an average of 8.06% and 5.94% on two scales of synthetic networks and real-world networks with real communities, and modularity is increased by an average of 0.80% on the real-world networks without real communities.

• KSII Transactions on Internet and Information Systems (TIIS)
• /
• 제14권4호
• /
• pp.1579-1602
• /
• 2020

In the design of production system, buffer capacity allocation is a major step. Through polymorphism analysis of production capacity and production capability, this paper investigates a buffer allocation optimization problem aiming at the multi-stage production line including unreliable machines, which is concerned with maximizing the system theoretical production rate and minimizing the system state entropy for a certain amount of buffers simultaneously. Stochastic process analysis is employed to establish Markov models for repairable modular machines. Considering the complex structure, an improved vector UGF (Universal Generating Function) technique and composition operators are introduced to construct the system model. Then the measures to assess the system's multi-state reliability and structural complexity are given. Based on system theoretical production rate and system state entropy, mathematical model for buffer capacity optimization is built and optimized by a specific genetic algorithm. The feasibility and effectiveness of the proposed method is verified by an application of an engine head production line.

• 지능정보연구
• /
• 제25권3호
• /
• pp.43-62
• /
• 2019

한때, 이상 탐지 분야는 특정 데이터로부터 도출한 기초 통계량을 기반으로 이상 유무를 판단하는 방법이 지배적이었다. 이와 같은 방법론이 가능했던 이유는 과거엔 데이터의 차원이 단순하여 고전적 통계 방법이 효과적으로 작용할 수 있었기 때문이다. 하지만 빅데이터 시대에 접어들며 데이터의 속성이 복잡하게 변화함에 따라 더는 기존의 방식으로 산업 전반에 발생하는 데이터를 정확하게 분석, 예측하기 어렵게 되었다. 따라서 기계 학습 방법을 접목한 SVM, Decision Tree와 같은 모형을 활용하게 되었다. 하지만 지도 학습 기반의 모형은 훈련 데이터의 이상과 정상의 클래스 수가 비슷할 때만 테스트 과정에서 정확한 예측을 할 수 있다는 특수성이 있고 산업에서 생성되는 데이터는 대부분 정답 클래스가 불균형하기에 지도 학습 모형을 적용할 경우, 항상 예측되는 결과의 타당성이 부족하다는 문제점이 있다. 이러한 단점을 극복하고자 현재는 클래스 분포에 영향을 받지 않는 비지도 학습 기반의 모델을 바탕으로 이상 탐지 모형을 구성하여 실제 산업에 적용하기 위해 시행착오를 거치고 있다. 본 연구는 이러한 추세에 발맞춰 적대적 생성 신경망을 활용하여 이상 탐지하는 방법을 제안하고자 한다. 시퀀스 데이터를 학습시키기 위해 적대적 생성 신경망의 구조를 LSTM으로 구성하고 생성자의 LSTM은 2개의 층으로 각각 32차원과 64차원의 은닉유닛으로 구성, 판별자의 LSTM은 64차원의 은닉유닛으로 구성된 1개의 층을 사용하였다. 기존 시퀀스 데이터의 이상 탐지 논문에서는 이상 점수를 도출하는 과정에서 판별자가 실제데이터일 확률의 엔트로피 값을 사용하지만 본 논문에서는 자질 매칭 기법을 활용한 함수로 변경하여 이상 점수를 도출하였다. 또한, 잠재 변수를 최적화하는 과정을 LSTM으로 구성하여 모델 성능을 향상시킬 수 있었다. 변형된 형태의 적대적 생성 모델은 오토인코더의 비해 모든 실험의 경우에서 정밀도가 우세하였고 정확도 측면에서는 대략 7% 정도 높음을 확인할 수 있었다.