Among the research of artificial neural networks, the most important problem is how to select the appropriate parameters for an artificial neural network. In this paper, a new evolutionary algorithm called region reproduction algorithm (RRA) is introduced to optimize the parameters of neural networks. The algorithm firstly generates some regions in space and then the offspring in the region is reproduced by the fitness in the superior regions. Because the algorithm is more concerned in the superior regions, it has more probability to find the optimal than traditional algorithms. Experiments for the Apple stock price data and Dell stock price data shows that our proposed RRA-NN model performed better than the traditional GA-NN model and can give much faster learning speed.
{"title":"A Region Reproduction Algorithm for Optimization of Neural Networks","authors":"Ya-ou Zhao, Yuehui Chen, Wei Li","doi":"10.1109/ICNC.2008.408","DOIUrl":"https://doi.org/10.1109/ICNC.2008.408","url":null,"abstract":"Among the research of artificial neural networks, the most important problem is how to select the appropriate parameters for an artificial neural network. In this paper, a new evolutionary algorithm called region reproduction algorithm (RRA) is introduced to optimize the parameters of neural networks. The algorithm firstly generates some regions in space and then the offspring in the region is reproduced by the fitness in the superior regions. Because the algorithm is more concerned in the superior regions, it has more probability to find the optimal than traditional algorithms. Experiments for the Apple stock price data and Dell stock price data shows that our proposed RRA-NN model performed better than the traditional GA-NN model and can give much faster learning speed.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"94 1","pages":"39-43"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75969815","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
The particle swarm optimization (PSO) algorithm with inertial weight is improved by adopting the sigmoid function and applied to structural damage identification in frame structures in this paper. The theoretical background on the improved PSO (IPSO) algorithm is introduced, a suite of four famous benchmark functions are then employed for evaluation on the IPSO performance. Some comparative studies on single and multi-damage of a 2-story rigid frame structure have been carried out. The illustrated results show that the IPSO algorithm can significantly improve the performance of the conventional PSO algorithm. IPSO can locate and identify structural damage with higher accuracy.
{"title":"An Improved PSO Algorithm and Its Application to Structural Damage Detection","authors":"Ling Yu, Zu-yong Wan","doi":"10.1109/ICNC.2008.224","DOIUrl":"https://doi.org/10.1109/ICNC.2008.224","url":null,"abstract":"The particle swarm optimization (PSO) algorithm with inertial weight is improved by adopting the sigmoid function and applied to structural damage identification in frame structures in this paper. The theoretical background on the improved PSO (IPSO) algorithm is introduced, a suite of four famous benchmark functions are then employed for evaluation on the IPSO performance. Some comparative studies on single and multi-damage of a 2-story rigid frame structure have been carried out. The illustrated results show that the IPSO algorithm can significantly improve the performance of the conventional PSO algorithm. IPSO can locate and identify structural damage with higher accuracy.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"15 1","pages":"423-427"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86884139","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Artificial immune system (AIS) is a recently developed evolutionary technique inspired by the theory or immunology. It is a powerful method for pattern recognition due to its self learning capability and memory acquisition. In this paper, we propose a complex artificial immune system (CAIS) based on an immune mechanism which emphasizes the antigen-presenting process during the immune response. The patterns encountered by the immune system are manipulated and memorized as complex styles. Compared with the traditional binary immune systems, the CAIS has a characteristic of invariant feature to recognize transformation,such as translation, scale or rotation ofpatterns. Simulation results based on arabic numerals show that the CAIS has a better acquisition of immune memory and noisy tolerance ability.
{"title":"A Complex Artificial Immune System","authors":"Wei Wang, Shangce Gao, Zheng Tang","doi":"10.1109/ICNC.2008.263","DOIUrl":"https://doi.org/10.1109/ICNC.2008.263","url":null,"abstract":"Artificial immune system (AIS) is a recently developed evolutionary technique inspired by the theory or immunology. It is a powerful method for pattern recognition due to its self learning capability and memory acquisition. In this paper, we propose a complex artificial immune system (CAIS) based on an immune mechanism which emphasizes the antigen-presenting process during the immune response. The patterns encountered by the immune system are manipulated and memorized as complex styles. Compared with the traditional binary immune systems, the CAIS has a characteristic of invariant feature to recognize transformation,such as translation, scale or rotation ofpatterns. Simulation results based on arabic numerals show that the CAIS has a better acquisition of immune memory and noisy tolerance ability.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"63 1","pages":"597-601"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86893161","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This paper points out some drawbacks and gives some modifications to the strong planning algorithm. In particular, a set of states is found by using of adjacency matrix that corresponds to a non-deterministic state-transition system. The set of states is composed of four parts, the first part is a part of initial states which can not reach goal states, if the first part is not empty, there is not strong planning; if a state which is not a initial state can not reach goal states,we put the state into the second part; the third part are the states that are unreachable from the initial states; so the station-action pairs which relate to the second part or third part are absolutely useless; the fourth part are the states which the initial states can not reach without passing any goal state before, the state-action pairs which relate to the fourth part are useless as well, because they move the execution away from the goal. So a great many of state-action pairs can be eliminated directly from the universal policy. Finally, the efficiency of the modified algorithm is illustrated by an example and some experiments.
{"title":"Improvements to the Strong Planning Algorithm","authors":"Zhonghua Wen, Qiwei Yang, Jinhua Zheng, Jiang Zhu","doi":"10.1109/ICNC.2008.723","DOIUrl":"https://doi.org/10.1109/ICNC.2008.723","url":null,"abstract":"This paper points out some drawbacks and gives some modifications to the strong planning algorithm. In particular, a set of states is found by using of adjacency matrix that corresponds to a non-deterministic state-transition system. The set of states is composed of four parts, the first part is a part of initial states which can not reach goal states, if the first part is not empty, there is not strong planning; if a state which is not a initial state can not reach goal states,we put the state into the second part; the third part are the states that are unreachable from the initial states; so the station-action pairs which relate to the second part or third part are absolutely useless; the fourth part are the states which the initial states can not reach without passing any goal state before, the state-action pairs which relate to the fourth part are useless as well, because they move the execution away from the goal. So a great many of state-action pairs can be eliminated directly from the universal policy. Finally, the efficiency of the modified algorithm is illustrated by an example and some experiments.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"85 1","pages":"506-510"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"87108328","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Nondeterminism in a stochastic process algebras (SPA) is resolved by a scheduler. For given scheduler D and a starting state, D induces a stochastic process MD on an SPA M. In this paper, by definition the path probability and using dynamic programming method of mathematic on a induced stochastic process algebras MD, we give an algorithm to compute the optimize path probability to reach a set of goal states within a time bound on SPA from the start state s0 and analysis the complexity of our algorithm.
{"title":"An Optimize Algorithm for Path Probabilities in Stochastic Process Algebras","authors":"Xiyng Zhao","doi":"10.1109/ICNC.2008.316","DOIUrl":"https://doi.org/10.1109/ICNC.2008.316","url":null,"abstract":"Nondeterminism in a stochastic process algebras (SPA) is resolved by a scheduler. For given scheduler D and a starting state, D induces a stochastic process MD on an SPA M. In this paper, by definition the path probability and using dynamic programming method of mathematic on a induced stochastic process algebras MD, we give an algorithm to compute the optimize path probability to reach a set of goal states within a time bound on SPA from the start state s0 and analysis the complexity of our algorithm.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"106 1","pages":"51-55"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"88043100","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
To address this problem of bounded rationality of resource consumers, an evolutionary algorithm for grid resource allocation is proposed in this paper. Firstly, the problem of multiple users competing for a common resource is formulated as a symmetric game. Secondly, replicated dynamic mechanism is used to produce evolutionary stable point that leads to a satisfied allocation scenario. Finally, the relationships among evolutionary stable point, valuation functions and convergence time are discussed in detail. The results of experiment show that the proposed evolutionary algorithm is convergent and generates better utility results compare to traditional algorithm.
{"title":"Grid Resource Allocation Based on Evolutionary Algorithm","authors":"Zhi-jie Li, Fei-xue Huang","doi":"10.1109/ICNC.2008.49","DOIUrl":"https://doi.org/10.1109/ICNC.2008.49","url":null,"abstract":"To address this problem of bounded rationality of resource consumers, an evolutionary algorithm for grid resource allocation is proposed in this paper. Firstly, the problem of multiple users competing for a common resource is formulated as a symmetric game. Secondly, replicated dynamic mechanism is used to produce evolutionary stable point that leads to a satisfied allocation scenario. Finally, the relationships among evolutionary stable point, valuation functions and convergence time are discussed in detail. The results of experiment show that the proposed evolutionary algorithm is convergent and generates better utility results compare to traditional algorithm.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"1 1","pages":"455-459"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86437481","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
This study discusses the pedagogical potential of an interactive response system system used in conjunction with interactive whiteboards technology to enhance learning in elementary school. Our empirical study examined the influence of interactivity on learning outcome and learner satisfaction in e-learning environments. Three different settings were studied: two were e-learning environments-with interactive response, and without interactive response. The third was the traditional classroom environment. Results of the experiment showed that the value of interactivity for learning effectiveness was contingent upon the provision of interactive response. Students in the e-learning environment that provided interactive response achieved significantly better learning performance and a higher level of learner satisfaction than those in other settings. However, students who used the e-learning environment that provided without interactive response did not improve either. The findings suggest that it may be important to integrate interactive response into e-learning systems.
{"title":"Using an Interactive Response System in Conjunction with Interactive Whiteboards Technology to Enhance Learning","authors":"Huan-Ming Chuang, Chia-Cheng Shen, Li-Chuan Wang","doi":"10.1109/ICNC.2008.40","DOIUrl":"https://doi.org/10.1109/ICNC.2008.40","url":null,"abstract":"This study discusses the pedagogical potential of an interactive response system system used in conjunction with interactive whiteboards technology to enhance learning in elementary school. Our empirical study examined the influence of interactivity on learning outcome and learner satisfaction in e-learning environments. Three different settings were studied: two were e-learning environments-with interactive response, and without interactive response. The third was the traditional classroom environment. Results of the experiment showed that the value of interactivity for learning effectiveness was contingent upon the provision of interactive response. Students in the e-learning environment that provided interactive response achieved significantly better learning performance and a higher level of learner satisfaction than those in other settings. However, students who used the e-learning environment that provided without interactive response did not improve either. The findings suggest that it may be important to integrate interactive response into e-learning systems.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"80 1","pages":"657-661"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86237154","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In order to obtain a reasonable method for new share pricing, new hybrid models based on self-organizing map with 2 winners (SOM2W) and radial basis function (RBF) neural network with characteristics of intelligence are proposed and applied to new share pricing in this paper. To enhance the dynamic competition and clustering capability of SOM2W network, and improve the precision of solutions, a tabu-mapping method is also used to avoid the same output node to be mapped by more than one input. Firstly, we use SOM2W model to clustering for stocks. The financial indexes reflecting the whole performance level of companies are used in the simulated experiments, so that the level of each stock can be confirmed. Then we use RBF neural network to simulate the system of the black box of stock to make a price for stocks. Experimental results show that the proposed hybrid models could provide a feasible approach and reference basis for new share pricing, which has potential applications in the financial field.
{"title":"SOM2W and RBF Neural Network-Based Hybrid Models and Their Applications to New Share Pricing","authors":"Xuming Han, Limin Wang, Xiaohu Shi, Yanchun Liang","doi":"10.1109/ICNC.2008.346","DOIUrl":"https://doi.org/10.1109/ICNC.2008.346","url":null,"abstract":"In order to obtain a reasonable method for new share pricing, new hybrid models based on self-organizing map with 2 winners (SOM2W) and radial basis function (RBF) neural network with characteristics of intelligence are proposed and applied to new share pricing in this paper. To enhance the dynamic competition and clustering capability of SOM2W network, and improve the precision of solutions, a tabu-mapping method is also used to avoid the same output node to be mapped by more than one input. Firstly, we use SOM2W model to clustering for stocks. The financial indexes reflecting the whole performance level of companies are used in the simulated experiments, so that the level of each stock can be confirmed. Then we use RBF neural network to simulate the system of the black box of stock to make a price for stocks. Experimental results show that the proposed hybrid models could provide a feasible approach and reference basis for new share pricing, which has potential applications in the financial field.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"106 1","pages":"538-542"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"86427461","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
Free-riding is a grave threat against the existence and efficient operation of Peer-to-peer (P2P) networks. This paper proposes a rank-based incentive mechanism to restrain the free-riding. In this mechanism, each user has a rank according to its utility value and it can just visit the corresponding rating files with its rank in the system. The utility function in this paper takes the absolute contribution value and the physical performance into account for the fairness and their value are calculated by Analytic Hierarchy Process. Besides, a pyramidal rank structure is constructed in this mechanism, and the computation formula for the nodepsilas transfer in each layer is established. The simulation verifies that this mechanism is effective and practical.
{"title":"Rank-Based Incentive Mechanism for Peer-to-Peer Networks","authors":"Yuhua Liu, Chun Yang, Kaihua Xu, Hongcai Chen","doi":"10.1109/ICNC.2008.641","DOIUrl":"https://doi.org/10.1109/ICNC.2008.641","url":null,"abstract":"Free-riding is a grave threat against the existence and efficient operation of Peer-to-peer (P2P) networks. This paper proposes a rank-based incentive mechanism to restrain the free-riding. In this mechanism, each user has a rank according to its utility value and it can just visit the corresponding rating files with its rank in the system. The utility function in this paper takes the absolute contribution value and the physical performance into account for the fairness and their value are calculated by Analytic Hierarchy Process. Besides, a pyramidal rank structure is constructed in this mechanism, and the computation formula for the nodepsilas transfer in each layer is established. The simulation verifies that this mechanism is effective and practical.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"76 1","pages":"252-256"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83847995","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
In chemistry, there are some algebra questions about equation roots that need to be solved. So far, there are some traditional methods such as graphic method and Newton iteration method and so on. As well known, the traditional methods have some shortcomings such as being easily affected by initial value and not being able to obtain very high precision. In this paper, evolution strategy algorithm is applied to solve the problems of complex chemistry equation roots by transforming them into the function optimization ones, making full use of such characteristics of the evolution strategy as being self-adaptive to gradually approach the optimal solution through optimized selection, recombination, and mutation. Finally, several computer simulation results show that the method in this paper has rapid convergence speed and powerful performance.
{"title":"Finding Roots of Complex Chemistry Equations Based on Evolution Strategies","authors":"Yongquan Zhou, Delong Guo","doi":"10.1109/ICNC.2008.178","DOIUrl":"https://doi.org/10.1109/ICNC.2008.178","url":null,"abstract":"In chemistry, there are some algebra questions about equation roots that need to be solved. So far, there are some traditional methods such as graphic method and Newton iteration method and so on. As well known, the traditional methods have some shortcomings such as being easily affected by initial value and not being able to obtain very high precision. In this paper, evolution strategy algorithm is applied to solve the problems of complex chemistry equation roots by transforming them into the function optimization ones, making full use of such characteristics of the evolution strategy as being self-adaptive to gradually approach the optimal solution through optimized selection, recombination, and mutation. Finally, several computer simulation results show that the method in this paper has rapid convergence speed and powerful performance.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"332 1","pages":"488-492"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"82943675","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":0,"RegionCategory":"","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}
京公网安备 11010802042870号
京ICP备2023020795号-1
扫码关注我们
求助内容：
应助结果提醒方式：