Application of financial information systems requires instant and fast response for continually changing market conditions. The purpose of this paper is to construct a mutual fund performance evaluation model utilizing the fast adaptive neural network classifier (FANNC), and to compare our results with those from a backpropagation neural networks (BPN) model. In our experiment, the FANNC approach requires much less time than the BPN approach to evaluate mutual fund performance. RMS is also superior for FANNC. These results hold for both classification problems and for prediction problems, making FANNC ideal for financial applications which require massive volumes of data and routine updates.
{"title":"Mutual Fund Performance Evaluation System Using Fast Adaptive Neural Network Classifier","authors":"Kehluh Wang, Szuwei Huang, Yi-Hsuan Chen","doi":"10.1109/ICNC.2008.756","DOIUrl":"https://doi.org/10.1109/ICNC.2008.756","url":null,"abstract":"Application of financial information systems requires instant and fast response for continually changing market conditions. The purpose of this paper is to construct a mutual fund performance evaluation model utilizing the fast adaptive neural network classifier (FANNC), and to compare our results with those from a backpropagation neural networks (BPN) model. In our experiment, the FANNC approach requires much less time than the BPN approach to evaluate mutual fund performance. RMS is also superior for FANNC. These results hold for both classification problems and for prediction problems, making FANNC ideal for financial applications which require massive volumes of data and routine updates.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75335220","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}
Recently, it becomes more important to ensure the quality of the products as copper strip manufacturing has been highly developed. The most difficult problem in process control and automatic inspection is classification of surface defects, so we develop an improved RBF (radial basis function) neural network classifier based on SVM (support vector machine) to automatically learn complicated defect patterns and use pseudo Zernike moment invariant as the defect feature. The optimal initial parameters of RBF network are gained through SVM, which has resolved the problems in traditional methods, e.g. long learning time, and easily getting into local minimum, etc. Furthermore, a BP learning algorithm is presented to adjust these hidden node parameters as well as the weights of the SVM-RBF. The experimental results show that the method is effective.
{"title":"Copper Strip Surface Defects Inspection Based on SVM-RBF","authors":"Ruiyu Liang, Yanqiong Ding, Xuewu Zhang, Jiasheng Chen","doi":"10.1109/ICNC.2008.271","DOIUrl":"https://doi.org/10.1109/ICNC.2008.271","url":null,"abstract":"Recently, it becomes more important to ensure the quality of the products as copper strip manufacturing has been highly developed. The most difficult problem in process control and automatic inspection is classification of surface defects, so we develop an improved RBF (radial basis function) neural network classifier based on SVM (support vector machine) to automatically learn complicated defect patterns and use pseudo Zernike moment invariant as the defect feature. The optimal initial parameters of RBF network are gained through SVM, which has resolved the problems in traditional methods, e.g. long learning time, and easily getting into local minimum, etc. Furthermore, a BP learning algorithm is presented to adjust these hidden node parameters as well as the weights of the SVM-RBF. The experimental results show that the method is effective.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75362384","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 this paper, the notion of vector-valued multiresolution analysis and biorthogonal vector-valued wavelets is introduced. The existence of compactly supported biorthogonal vector-valued wavelets associated with a pair of biorthogonal compactly supported vector-valued scaling functions is investigated. An algorithm for constructing a class of biorthogonal compactly supported vector-valued wavelet functions is presented by using multiresolution analysis and matrix theory.
{"title":"Construction of Biorthogonal Compactly Supported Vector-Valued Wavelets","authors":"Tongqi Zhang","doi":"10.1109/ICNC.2008.180","DOIUrl":"https://doi.org/10.1109/ICNC.2008.180","url":null,"abstract":"In this paper, the notion of vector-valued multiresolution analysis and biorthogonal vector-valued wavelets is introduced. The existence of compactly supported biorthogonal vector-valued wavelets associated with a pair of biorthogonal compactly supported vector-valued scaling functions is investigated. An algorithm for constructing a class of biorthogonal compactly supported vector-valued wavelet functions is presented by using multiresolution analysis and matrix theory.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"75529262","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 aim of this research is to investigate the characterization of firing pattern of neuron at CA1 for dysfunctional memory mice via entropy coding. The spike trains were recorded at CA1 of hippocampus slice for two groups: senescence-accelerated-prone (SAM) -P/8 type (SAM-P/8) mice group and normal control mice group. Shannon entropy based on inter spike intervals (ISIs) histogram was used to measure the code of neuron activity at CA1 of hippocampus slice for two groups (10 samples for each group). The different of the entropy codes for two groups was tested by t-test. The results show that Shannon entropy forSAM-P/8 mice group was 9.30plusmn0.44 bit, which is apparently greater than that for normal mice group, which was 7.26plusmn0.33 bit.The conclusion is that the higher entropy value for SAM-P/8 mice group is revealed lower information level than the normal group, which suggests the dysfunction of synaptic plasticity for senescence-accelerated-prone mice. The results might support the research of memory dysfunction from the view of neural coding pattern.
本研究的目的是通过熵编码研究功能障碍小鼠CA1神经元放电模式的特征。在衰老加速倾向(SAM) -P/8型(SAM-P/8)小鼠组和正常对照小鼠两组海马CA1区记录峰列。采用基于峰间间隔(ISIs)直方图的Shannon熵测量两组(每组10个样本)海马切片CA1神经元活动编码。用t检验检验两组间熵码的差异。结果表明，sam - p /8小鼠组的Shannon熵为9.30plusmn0.44 bit，明显大于正常小鼠组的7.26plusmn0.33 bit。综上所述，SAM-P/8小鼠高熵值组的信息水平低于正常组，提示衰老加速小鼠突触可塑性功能障碍。这一结果可能从神经编码模式的角度支持记忆功能障碍的研究。
{"title":"Entropy Coding of Neuron Firings at Hippocampus CA1 for Memory Dysfunctional Mice","authors":"Xiaoping Zheng, Xian Tian, Tiaotiao Liu, H. Tao","doi":"10.1109/ICNC.2008.441","DOIUrl":"https://doi.org/10.1109/ICNC.2008.441","url":null,"abstract":"The aim of this research is to investigate the characterization of firing pattern of neuron at CA1 for dysfunctional memory mice via entropy coding. The spike trains were recorded at CA1 of hippocampus slice for two groups: senescence-accelerated-prone (SAM) -P/8 type (SAM-P/8) mice group and normal control mice group. Shannon entropy based on inter spike intervals (ISIs) histogram was used to measure the code of neuron activity at CA1 of hippocampus slice for two groups (10 samples for each group). The different of the entropy codes for two groups was tested by t-test. The results show that Shannon entropy forSAM-P/8 mice group was 9.30plusmn0.44 bit, which is apparently greater than that for normal mice group, which was 7.26plusmn0.33 bit.The conclusion is that the higher entropy value for SAM-P/8 mice group is revealed lower information level than the normal group, which suggests the dysfunction of synaptic plasticity for senescence-accelerated-prone mice. The results might support the research of memory dysfunction from the view of neural coding pattern.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74224312","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 introduced a new type of fault testing device for underground electric heating cable. A new design concept was adopted and intelligent control techniques were used in the trouble-shooting testing system for buried electric cable. The testing principles of the short-circuit fault point and the breaking point were discussed. The performance and functions of the device were described.
{"title":"A Device for Fault Testing of the Underground Electric Heating Cable","authors":"Bing Li, Yilin Shen, Li Li","doi":"10.1109/ICNC.2008.580","DOIUrl":"https://doi.org/10.1109/ICNC.2008.580","url":null,"abstract":"This paper introduced a new type of fault testing device for underground electric heating cable. A new design concept was adopted and intelligent control techniques were used in the trouble-shooting testing system for buried electric cable. The testing principles of the short-circuit fault point and the breaking point were discussed. The performance and functions of the device were described.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74249506","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 ensure fast adaptation and security of social and computerized systems to changing environments, targets of perceptron based classifiers ought to vary during training process. To determine optimal differences between target values (stimulation, arousal) we suggest using genetically evolving multi-agent systems aimed to extract necessary information from sequences of the changes. A specially designed additional feedback chain allows updating the target values faster.
{"title":"A Target Value Control While Training the Perceptrons in Changing Environments","authors":"S. Raudys","doi":"10.1109/ICNC.2008.891","DOIUrl":"https://doi.org/10.1109/ICNC.2008.891","url":null,"abstract":"To ensure fast adaptation and security of social and computerized systems to changing environments, targets of perceptron based classifiers ought to vary during training process. To determine optimal differences between target values (stimulation, arousal) we suggest using genetically evolving multi-agent systems aimed to extract necessary information from sequences of the changes. A specially designed additional feedback chain allows updating the target values faster.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"74776153","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}
Chen Feng, Guangrong Ji, Junna Cheng, Xuefeng Liu, Jie Zhang
This paper presents an improved method for estimating the local fractal dimension. The bright points in the local fractal dimension (LFD) map image derived from previous method may influence the effect of edge detection. To solve this problem, we use 0 as the replaced value to supplement the values which do not exist. At the same time, based on the improved local fractal dimension we have furthermore proposed an edge detection method according to the difference between two LFD map images which are computed using local windows of different sizes. The experimental results have proved the effectiveness of our method.
{"title":"Image Edge Detection Based on Improved Local Fractal Dimension","authors":"Chen Feng, Guangrong Ji, Junna Cheng, Xuefeng Liu, Jie Zhang","doi":"10.1109/ICNC.2008.447","DOIUrl":"https://doi.org/10.1109/ICNC.2008.447","url":null,"abstract":"This paper presents an improved method for estimating the local fractal dimension. The bright points in the local fractal dimension (LFD) map image derived from previous method may influence the effect of edge detection. To solve this problem, we use 0 as the replaced value to supplement the values which do not exist. At the same time, based on the improved local fractal dimension we have furthermore proposed an edge detection method according to the difference between two LFD map images which are computed using local windows of different sizes. The experimental results have proved the effectiveness of our method.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"72695229","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}
Shun Long, Hui-Jin Wang, Jian-Hua Cai, Changhui Liu
Learning is an integral part of an intelligent agent's problem-solving ability. Most research works on learning assume that information are provided in a complete and well-structured manner and/or when rules are clearly specified, therefore have difficulties in handling cases with incomplete and unstructured data. This paper presents a novel approach to deal with this difficulty. It uses information retrieval techniques to enhance an agent's reasoning and problem-solving ability when only incomplete and unstructured information are available. Preliminary experimental results show that the integration of information retrieval can effectively help an agent to analyze the problem before solving it.
{"title":"Enhancing Intelligent Agents with Information Retrieval Techniques","authors":"Shun Long, Hui-Jin Wang, Jian-Hua Cai, Changhui Liu","doi":"10.1109/ICNC.2008.15","DOIUrl":"https://doi.org/10.1109/ICNC.2008.15","url":null,"abstract":"Learning is an integral part of an intelligent agent's problem-solving ability. Most research works on learning assume that information are provided in a complete and well-structured manner and/or when rules are clearly specified, therefore have difficulties in handling cases with incomplete and unstructured data. This paper presents a novel approach to deal with this difficulty. It uses information retrieval techniques to enhance an agent's reasoning and problem-solving ability when only incomplete and unstructured information are available. Preliminary experimental results show that the integration of information retrieval can effectively help an agent to analyze the problem before solving it.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77614444","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 this paper, a visual attention model for dynamic scenes based on two-pathway processing in brain is proposed. A saliency map is created to measure the static features, and a motion conspicuity map is created to measure the dynamic feature. Both maps are integrated in an IFNN which is employed to simulate the interactive aspects of two-pathway processing in brain. Selection of attention is accomplished by the interspike interval of the neural network. The experiments results show the effective and efficient performance of the model.
{"title":"A Visual Attention Model for Dynamic Scenes Based on Two-Pathway Processing in Brain","authors":"Jiawei Chen, Changle Zhou, Kunhui Lin, Yanyun Qu","doi":"10.1109/ICNC.2008.392","DOIUrl":"https://doi.org/10.1109/ICNC.2008.392","url":null,"abstract":"In this paper, a visual attention model for dynamic scenes based on two-pathway processing in brain is proposed. A saliency map is created to measure the static features, and a motion conspicuity map is created to measure the dynamic feature. Both maps are integrated in an IFNN which is employed to simulate the interactive aspects of two-pathway processing in brain. Selection of attention is accomplished by the interspike interval of the neural network. The experiments results show the effective and efficient performance of the model.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"77859743","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 this paper a new polynomial interior-point algorithm for monotone mixed linear complementarity problem is presented. The algorithm is based on a new technique for finding a class of search directions and the strategy of the central path. At each iteration, we use only full-Newton step. Moreover, we obtain the currently best known iteration bound for the algorithm with small-update method, namely,O(radic(n log (nisin))), which is as good as the linear analogue.
{"title":"A New Polynomial Interior-Point Algorithm for Monotone Mixed Linear Complementarity Problem","authors":"Guoqiang Wang, Xinzhong Cai, Y. Yue","doi":"10.1109/ICNC.2008.245","DOIUrl":"https://doi.org/10.1109/ICNC.2008.245","url":null,"abstract":"In this paper a new polynomial interior-point algorithm for monotone mixed linear complementarity problem is presented. The algorithm is based on a new technique for finding a class of search directions and the strategy of the central path. At each iteration, we use only full-Newton step. Moreover, we obtain the currently best known iteration bound for the algorithm with small-update method, namely,O(radic(n log (nisin))), which is as good as the linear analogue.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":null,"pages":null},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79834851","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：