http://chineseinput.net/에서 pinyin(병음)방식으로 중국어를 변환할 수 있습니다.
변환된 중국어를 복사하여 사용하시면 됩니다.
Analysis of Fuzzy Class Association Rule Mining Based on Genetic Network Programming
Ci Chen,Shingo Mabu,Chuan Yue,Kaoru Shimada,Kotaro Hirasawa 제어로봇시스템학회 2009 제어로봇시스템학회 국제학술대회 논문집 Vol.2009 No.8
Classification rule mining is apractical data mining technique widely used inreal world. In the previous work, we have put for ward a fuzzy classassociation rule mining method based on Genetic Network Programming and applied it to network in trusion detection system which proved it sefficiency and advantage. In this paper, a detailed comparison not only between fuzzy class association rule minings(FCARMs) with fixed fuzzy membership functions and with evolved fuzzy membership functions, but also between FCARMs with and with out probability no detransition are carried out. The aim of this paper is to provide experimental analysis on the characteristics of FCARMs with different implementations. Experimental results conducted on real world database, KDD99 Cup and DAPRA 98 database from MITL incoln Laboratory, are studied to verify the comparison.
Genetic Network Programming with General Individual Reconstruction
Fengming Ye,Shingo Mabu,Lutao Wang,Kotaro Hirasawa 제어로봇시스템학회 2009 제어로봇시스템학회 국제학술대회 논문집 Vol.2009 No.8
Genetic Network Programming(GNP) which has been developed for dealing with problems in dynamic envi-ronments is a newly propose devolutionary approach with the data structure of directed graphs. GNP has been used in many different areas such as datamining, extracting trading rules of stock markets, elevator supervised control systems, etc and has obtained some out standing results. Focusing on GNP’s distinguishing expressionability of the graph structure, this paper proposes a method named Genetic Network Program-ming with General Individual Recon struction(GNP with GIR) which reconstructs the gene of randomly selected individuals and then under goes the special genetic operations by using the transition information of betterin dividuals. The unique indi-vidual reconstruction and genetic operations make individuals not only learn the experiences of better individuals but also strength enexploratio and exploration ability. GNP with GIR will be applied to the tile-world which is an excellent bench mark for evaluating the proposed architecture. The performances of GNP with GIR will becompared with conventional GNP demonstrating its superiority.
Class Association Rule Mining with Correlation Measuresusing Genetic Network Programming
Eloy Gonzales,Shingo Mabu,Karla Taboada,Kaoru Shimada,Kotaro Hirasawa 제어로봇시스템학회 2009 제어로봇시스템학회 국제학술대회 논문집 Vol.2009 No.8
Association rule mining is one of the tasks of datamining and it has been extensively studied recently. As a consequence, several methods for extractin gas sociation rules have been developed during the last years. Most of the muse the supportand confidence framework to extract the association rules. Researches are able to extract strong rules using this framework. However these measures are not good enough to solve the quality problems of the rules. A new data mining method using Genetic Network Programming(GNP) has also been developed recently which uses the χ<SUP>2</SUP> (chi-squared) as a correlation measure and its effectiveness has been shown for different data sets[1][2]. To enhance the correlation degree and comprehensibility of association rule, several correlation measures including lift, χ<SUP>2</SUP>, all-confidence and cosineare studied in this paper when they are in corporated in the conventional GNP based mining algorithm. A comparison between the correlation measures is made in the simulations when they are incorporated separately in to the GNP based mining method. Finally, the association rules extracted using different correlation measures are applied to the classification problems and the prediction accuracies of them are evaluated.
Agent Bidding Strategy of Multiple Round English Auction based on Genetic Network Programming
Chuan Yue,Shingo Mabu,Yan Chen,Yu Wang,Kotaro Hirasawa 제어로봇시스템학회 2009 제어로봇시스템학회 국제학술대회 논문집 Vol.2009 No.8
The auction mechanism is widely used in web-based sites and originally designed for human beings, but it might not be the most efficientonein the future, while, there is ademand of evolutionary computation auction agents adaptable to the dynamic auction environments. In this paper, we have applied Genetic Network Programming(GNP) to auction agents to determine a bid at each time step and developed multiple round English Auction mechanisms based on multi-agent systems. In the simulations, we provide comparisons of the proposed method with existing ones. As a result, it has been found that the proposed method could help agents to evolve their strategies generation by generation to get more goods with less money. Also, GNP has a good performance of helping the agent to find out the suitable strategy under the current situation.
Adaptive Controller for Double-Deck Elevator Systemusing Genetic Network Programming
Johanna Mansilla,Shingo Mabu,Lu Yu,Kotaro Hirasawa 제어로봇시스템학회 2009 제어로봇시스템학회 국제학술대회 논문집 Vol.2009 No.8
In this paper, an improved approach is proposed based on an updating strategy using Genetic Network Programming(GNP) for the controller of Double-Deck Elevator System(DDES). Since the elevator controller has to deal with constant changes of its environment, our approach is proposed to deal with better the environment changes, resulting in a reduction of the waiting time and increasing the transportation capacity. This updating looks for ward to contributing to the efficient adjustment of the system periodically according to the gained system information. The performance of the proposed method is evaluated by comparis on with the conventional GNP method, which does not update the controller. By this evaluation, the enhancement of our model is confirmed.
Guangfei Yang,Shingo Mabu,Kaoru Shimada,Kotaro Hirasawa 제어로봇시스템학회 2009 제어로봇시스템학회 국제학술대회 논문집 Vol.2009 No.8
In this paper, we propose an evolutionary approach to rank association rules for classification. The association rules are ranked by their support, confidence and length in one of the most important associative classification method, Classification based on Multiple Association Rule(CMAR). However, from some empirical studies, we find that if the rules are ranked by some equations first, the classification accuracy will be improved in some data sets. In order to generate such equations effectively, we propose a Rule Rank model based on Genetic Network Programming(GNP). The experimental results show that our method could improve the classification accuracies effectively.
Global Optimal Routing for Traffic Systems with Multiple ODs using Genetic Algorithm
Yu Wang,Shingo Mabu,Chuan Yue,Manoj Kanta Mainali,Kotaro Hirasawa 제어로봇시스템학회 2009 제어로봇시스템학회 국제학술대회 논문집 Vol.2009 No.8
The multiple origins multiple destinations routing(MOMDR) problem becomes extremely complicated when considering the traffic volumes on road sections. When solving this kind of problem, only heuristical gorithms have practical values because it is a typical NP-Hard problem. This paper applies Genetic Algorithm(GA) to enhance Sorting-Randomizing-Adjusting-Updating(SRAU) algorithm[1]. The former papers hows that different processing order softhe origin-destinations(ODs) result indifferent solutions with different performances. Therefore, a heuristical gorithm for finding the best processing order of ODs can optimize SRAU algorithm. In this paper, every processing order of ODs is transformed into agene/chromo some of the individuals of GA; then the best gene can befound during the evolution of GA; finally, the best gene is transformed back to find the optimal solution of the problem. Sufficient simulations show that the proposed algorithm is more efficient than original SRAU algorithm. Also the consideration of the traffic volumes on the road sections enables the proposed method to apply to real traffic systems.
Karla Taboada,Shingo Mabu,Eloy Gonzales,Kaoru Shimada,Kotaro Hirasawa 제어로봇시스템학회 2009 제어로봇시스템학회 국제학술대회 논문집 Vol.2009 No.8
One of the most important issues in any association rule mining is the interpretation and evaluation of discov-eredrules. Thus, most algorithms employ the support-confidence framework for evaluating association and classification rules. Unfortunately, recent studies show that the supportand confidence measures are insufficient for filtering outun in-teresting association rules, for instance, even strong association rules can be uninteresting and misleading. To deal with this limitation, the support-confidence framework can be suplemented with additional interestingness measures based on statistical significance and correlation analysis. In this paper, a novel fuzzy association rule-based classification approach is proposed, where χ<SUP>2</SUP> is applied as a correlation measure. The algorithm is based on Genetic Network Programming(GNP) and discover comprehensible fuzzy association rules potentially useful for classification. GNP is an evolutionary optimization algorithm that uses directed graph structures as genes instead of strings and trees of Genetic Algorithms(GA) and Genetic Programming(GP), respectively. This feature contributes to creating quite compact programs and im-plicitly memorizing pastaction sequences. The proposed model consists of two major phases: 1) generating fuzzy class association rules by using GNP, 2) building a classifier based on the extracted fuzzy rules. In the first phase, χ<SUP>2</SUP> is used for computing the correlation of the rules to be integrated into the classifier. In the second phase, the χ<SUP>2</SUP> value is used as a weight of the rule when calculating the matching degree of the rule with new data. The performance of the proposed algorithm has been compared with other relevant algorithms and the experimental results have shown the advantages an effectiveness of the proposed model.
Yunlu Gong,Shingo Mabu,Ci Chen,Yifei Wang,Kotaro Hirasawa 제어로봇시스템학회 2009 제어로봇시스템학회 국제학술대회 논문집 Vol.2009 No.8
In this paper, a class association rule mining approach based on Genetic Network Programming(GNP) for detecting network intrusion combining misuse detection and anomaly detection is proposed. The proposed approach is an extension of the intrusion detection approach using GNP, so it can detectand distinguish normal, known in trusion and unknown intrusion. The simulation result shows that the detection rate is improved compared with traditional intrusion detection approach, and normal, known intrusion and unknown intrusion are distinguished with high accuracy.