Pub Date : 2005-10-13DOI: 10.1109/ICNNB.2005.1614713
Liu Fuyan
In this paper we propose an attribute selection approach, which is based on rough sets theory. The main feature of this method is that it not only takes the dependency degree of decision attributes on condition attributes into account, but also considers decision makers' priori knowledge about importance of condition attributes to decision attributes. It combines these two factors as a criterion of attribute selection. In addition, it uses a compound weights algorithm to implement a proper reduct. As a result, the most effective attribute subset is obtained, and a practical, reduced knowledge rule set can be acquired. In order to judge the effectiveness of the proposed approach, the knowledge rule set acquired is applied to a prototype simulation system of a part assembly cell for optimum control. Experimental results indicate that the attribute and reduct selection approach is more effective
{"title":"An Attribute Selection Approach and Its Application","authors":"Liu Fuyan","doi":"10.1109/ICNNB.2005.1614713","DOIUrl":"https://doi.org/10.1109/ICNNB.2005.1614713","url":null,"abstract":"In this paper we propose an attribute selection approach, which is based on rough sets theory. The main feature of this method is that it not only takes the dependency degree of decision attributes on condition attributes into account, but also considers decision makers' priori knowledge about importance of condition attributes to decision attributes. It combines these two factors as a criterion of attribute selection. In addition, it uses a compound weights algorithm to implement a proper reduct. As a result, the most effective attribute subset is obtained, and a practical, reduced knowledge rule set can be acquired. In order to judge the effectiveness of the proposed approach, the knowledge rule set acquired is applied to a prototype simulation system of a part assembly cell for optimum control. Experimental results indicate that the attribute and reduct selection approach is more effective","PeriodicalId":145719,"journal":{"name":"2005 International Conference on Neural Networks and Brain","volume":"8 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"128013145","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}
Pub Date : 2005-10-13DOI: 10.1109/ICNNB.2005.1615013
F. Jiang, A. P. Preethy, Yanqing Zhang
The properties of training data set such as size, distribution and number of attributes significantly contribute to the generalization error of a learning machine. A data set not well-distributed is prone to lead to a model with partial overfitting. The approach proposed in this paper for the binary classification enhances the useful data information by mining negative data based on the understanding of Chinese traditional Yin-Yang theory
{"title":"Compensating Hypothesis by Negative Data","authors":"F. Jiang, A. P. Preethy, Yanqing Zhang","doi":"10.1109/ICNNB.2005.1615013","DOIUrl":"https://doi.org/10.1109/ICNNB.2005.1615013","url":null,"abstract":"The properties of training data set such as size, distribution and number of attributes significantly contribute to the generalization error of a learning machine. A data set not well-distributed is prone to lead to a model with partial overfitting. The approach proposed in this paper for the binary classification enhances the useful data information by mining negative data based on the understanding of Chinese traditional Yin-Yang theory","PeriodicalId":145719,"journal":{"name":"2005 International Conference on Neural Networks and Brain","volume":"2 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132620205","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}
Pub Date : 2005-10-13DOI: 10.1109/ICNNB.2005.1614972
Guang Ye, Chen Guo
Based on simulated annealing (SA) and reinforcement learning (RL) algorithm, a hybrid intelligent controller is proposed to ship steering. The SA algorithm is a powerful way to solve hard combinatorial optimization problems, which is used to adjust the parameters of the controller in this paper. The RL algorithm shows its particular superiority in ship steering, which just needs simple fuzzy information. With the advantages of the two algorithms, the controller can overcome the influence of the wind, wave and flow, the limitation that data are not exactly accurate. At last, the results of the simulation show that the ship course can be properly controlled when changeable wind, wave, and measure error exists
{"title":"SA-RL Algorithm Based Ship Steering Controller","authors":"Guang Ye, Chen Guo","doi":"10.1109/ICNNB.2005.1614972","DOIUrl":"https://doi.org/10.1109/ICNNB.2005.1614972","url":null,"abstract":"Based on simulated annealing (SA) and reinforcement learning (RL) algorithm, a hybrid intelligent controller is proposed to ship steering. The SA algorithm is a powerful way to solve hard combinatorial optimization problems, which is used to adjust the parameters of the controller in this paper. The RL algorithm shows its particular superiority in ship steering, which just needs simple fuzzy information. With the advantages of the two algorithms, the controller can overcome the influence of the wind, wave and flow, the limitation that data are not exactly accurate. At last, the results of the simulation show that the ship course can be properly controlled when changeable wind, wave, and measure error exists","PeriodicalId":145719,"journal":{"name":"2005 International Conference on Neural Networks and Brain","volume":"40 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134023347","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}
Pub Date : 2005-10-13DOI: 10.1109/ICNNB.2005.1614681
Miao Kang, D. Palmer-Brown
ADFUNN is based on a linear piecewise neuron activation function that is modified by a novel gradient descent supervised learning algorithm. It has been applied to some linearly inseparable problems: XOR, Iris dataset, phrase recognition problem. In all cases it exhibited impressive generalisation classification ability with no hidden nodes. In addition, the learned functions support intelligent data analysis. In this paper, we improve the general learning rule of ADFUNN by using proximal proportionality to adapt neural activation functions more accurately. The learned functions are then smoothed in preparation for recognising their closest fit to analytical functions. We compare two different algorithms for smoothing the learned function curves: the simple moving average and least-squares polynomial smoothing. The smoothed curves prove to be accurate replacements for the natural language phrase recognition test case
{"title":"An adaptive function neural network (ADFUNN) classifier","authors":"Miao Kang, D. Palmer-Brown","doi":"10.1109/ICNNB.2005.1614681","DOIUrl":"https://doi.org/10.1109/ICNNB.2005.1614681","url":null,"abstract":"ADFUNN is based on a linear piecewise neuron activation function that is modified by a novel gradient descent supervised learning algorithm. It has been applied to some linearly inseparable problems: XOR, Iris dataset, phrase recognition problem. In all cases it exhibited impressive generalisation classification ability with no hidden nodes. In addition, the learned functions support intelligent data analysis. In this paper, we improve the general learning rule of ADFUNN by using proximal proportionality to adapt neural activation functions more accurately. The learned functions are then smoothed in preparation for recognising their closest fit to analytical functions. We compare two different algorithms for smoothing the learned function curves: the simple moving average and least-squares polynomial smoothing. The smoothed curves prove to be accurate replacements for the natural language phrase recognition test case","PeriodicalId":145719,"journal":{"name":"2005 International Conference on Neural Networks and Brain","volume":"96 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"132221423","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}
Pub Date : 2005-10-13DOI: 10.1109/ICNNB.2005.1614915
Wang Shoujue, Chen Xu, Liu Weijun
In this paper, we propose a new scheme for omnidirectional object-recognition in free space. The proposed scheme divides above problem into several omnidirectional object-recognition with different depression angles. An omnidirectional object-recognition system with oblique observation directions based on a new recognition theory-biomimetic pattern recognition (BPR) is discussed in detail. Based on it, we can get the size of training samples in the omnidirectional object-recognition system in free space. Omnidirectionally cognitive tests were done on various kinds of animal models of rather similar shapes. For the total 8400 tests, the correct recognition rate is 99.89%. The rejection rate is 0.11% and on the condition of zero error rates. Experimental results are presented to show that the proposed approach outperforms three types of SVMs with either a three degree polynomial kernel or a radial basis function kernel
{"title":"Object-Recognition with oblique observation directions Based on Biomimetic Pattern Recognition","authors":"Wang Shoujue, Chen Xu, Liu Weijun","doi":"10.1109/ICNNB.2005.1614915","DOIUrl":"https://doi.org/10.1109/ICNNB.2005.1614915","url":null,"abstract":"In this paper, we propose a new scheme for omnidirectional object-recognition in free space. The proposed scheme divides above problem into several omnidirectional object-recognition with different depression angles. An omnidirectional object-recognition system with oblique observation directions based on a new recognition theory-biomimetic pattern recognition (BPR) is discussed in detail. Based on it, we can get the size of training samples in the omnidirectional object-recognition system in free space. Omnidirectionally cognitive tests were done on various kinds of animal models of rather similar shapes. For the total 8400 tests, the correct recognition rate is 99.89%. The rejection rate is 0.11% and on the condition of zero error rates. Experimental results are presented to show that the proposed approach outperforms three types of SVMs with either a three degree polynomial kernel or a radial basis function kernel","PeriodicalId":145719,"journal":{"name":"2005 International Conference on Neural Networks and Brain","volume":"38 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134284956","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}
Pub Date : 2005-10-13DOI: 10.1109/ICNNB.2005.1614587
Decai Huang, Haidong Guo, Neng Qian
A hybrid genetic algorithm is presented for minimizing the range of lateness and make-span on parallel non-identical machines in this paper, and a dynamic fitness function is introduced too. The coding method of the hybrid genetic algorithm (HGA) is very simple because it utilized the property of effective optimal algorithm for solving the corresponding single machine problem. It made the implement of HGA be very easy. Numerical simulations illustrate that the HGA has the property of fast convergence, and can be used to solve larger size problems
{"title":"Hybrid Genetic Algorithm for Minimizing the Range of Lateness and Make-span on Non-identical Parallel Machines","authors":"Decai Huang, Haidong Guo, Neng Qian","doi":"10.1109/ICNNB.2005.1614587","DOIUrl":"https://doi.org/10.1109/ICNNB.2005.1614587","url":null,"abstract":"A hybrid genetic algorithm is presented for minimizing the range of lateness and make-span on parallel non-identical machines in this paper, and a dynamic fitness function is introduced too. The coding method of the hybrid genetic algorithm (HGA) is very simple because it utilized the property of effective optimal algorithm for solving the corresponding single machine problem. It made the implement of HGA be very easy. Numerical simulations illustrate that the HGA has the property of fast convergence, and can be used to solve larger size problems","PeriodicalId":145719,"journal":{"name":"2005 International Conference on Neural Networks and Brain","volume":"112 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134264490","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}
Pub Date : 2005-10-13DOI: 10.1109/ICNNB.2005.1614646
Yumin Zhang, Lingyao Wu, Lei Guo
In many practical processes, the measured information is the stochastic distribution of the system output rather than its value. In this paper the fault diagnosis (FD) problem is considered by using the output stochastic distributions. A multi-layer perceptron (MLP) neural network is adopted to approximate the probability density function (PDF) of the system outputs and nonlinear principal component analysis (NLPCA) is applied to reduce the model order for a lower-order model. For such a discrete-time dynamic model with nonlinearities, uncertainties and time delays, the concerned FD problem is investigated. The measure of estimation errors represented by the distances between two output PDFs, would be optimized to find the diagnosis filter gain. Simulation example is given for the weighting dynamics to demonstrate the effectiveness
{"title":"Filtering-Based Actuator Fault Diagnosis using MLP Neural Network for PDFs","authors":"Yumin Zhang, Lingyao Wu, Lei Guo","doi":"10.1109/ICNNB.2005.1614646","DOIUrl":"https://doi.org/10.1109/ICNNB.2005.1614646","url":null,"abstract":"In many practical processes, the measured information is the stochastic distribution of the system output rather than its value. In this paper the fault diagnosis (FD) problem is considered by using the output stochastic distributions. A multi-layer perceptron (MLP) neural network is adopted to approximate the probability density function (PDF) of the system outputs and nonlinear principal component analysis (NLPCA) is applied to reduce the model order for a lower-order model. For such a discrete-time dynamic model with nonlinearities, uncertainties and time delays, the concerned FD problem is investigated. The measure of estimation errors represented by the distances between two output PDFs, would be optimized to find the diagnosis filter gain. Simulation example is given for the weighting dynamics to demonstrate the effectiveness","PeriodicalId":145719,"journal":{"name":"2005 International Conference on Neural Networks and Brain","volume":"59 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134497624","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}
Pub Date : 2005-10-13DOI: 10.1109/ICNNB.2005.1614868
Zhongzhi Shi, Qingyong Li, Zheng Zheng
Perceptual learning should be considered as an active process that embeds particular abstraction, reformulation and approximation within the Abstraction framework. In this paper we focus on sparse coding theory and granular computing model for visual perceptual learning. We propose a novel sparse coding model, called here classification-oriented sparse coding (COSC) model for learning sparse and informative structures in natural images for visual classification task, combining the discriminability constraint supervised by visual classification task, besides the sparseness criteria. An attention-guided sparse coding model will be also proposed in the paper. This model is a data-driven attention module based on the response saliency. For the granular computing based on tolerance relation we construct a more uniform granulation model, which is established on both consecutive space and discrete attribute space.
{"title":"Visual Perceptual Learning","authors":"Zhongzhi Shi, Qingyong Li, Zheng Zheng","doi":"10.1109/ICNNB.2005.1614868","DOIUrl":"https://doi.org/10.1109/ICNNB.2005.1614868","url":null,"abstract":"Perceptual learning should be considered as an active process that embeds particular abstraction, reformulation and approximation within the Abstraction framework. In this paper we focus on sparse coding theory and granular computing model for visual perceptual learning. We propose a novel sparse coding model, called here classification-oriented sparse coding (COSC) model for learning sparse and informative structures in natural images for visual classification task, combining the discriminability constraint supervised by visual classification task, besides the sparseness criteria. An attention-guided sparse coding model will be also proposed in the paper. This model is a data-driven attention module based on the response saliency. For the granular computing based on tolerance relation we construct a more uniform granulation model, which is established on both consecutive space and discrete attribute space.","PeriodicalId":145719,"journal":{"name":"2005 International Conference on Neural Networks and Brain","volume":"62 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"134517713","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}
Pub Date : 2005-10-13DOI: 10.1109/ICNNB.2005.1615005
Xiaoli Li, Tong Wang, Zhenlong Du
A novel method of audio retrieval based on tolerant rough set theory is proposed. The existence of noise and audio characteristics affect the exactness of retrieval set generated by conventional approaches. In this paper, we construct audio feature set by audio features, retrieve and match the audio clip in the approximate space of tolerances rough set. Experiments show that our method overcomes the limitation of equivalent rough set in audio retrieval, and improves the retrieval efficiency
{"title":"Audio Retrieval Based on Tolerance Rough Sets","authors":"Xiaoli Li, Tong Wang, Zhenlong Du","doi":"10.1109/ICNNB.2005.1615005","DOIUrl":"https://doi.org/10.1109/ICNNB.2005.1615005","url":null,"abstract":"A novel method of audio retrieval based on tolerant rough set theory is proposed. The existence of noise and audio characteristics affect the exactness of retrieval set generated by conventional approaches. In this paper, we construct audio feature set by audio features, retrieve and match the audio clip in the approximate space of tolerances rough set. Experiments show that our method overcomes the limitation of equivalent rough set in audio retrieval, and improves the retrieval efficiency","PeriodicalId":145719,"journal":{"name":"2005 International Conference on Neural Networks and Brain","volume":"102 2","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"114025445","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 qualitative mapping (QM) model for judging a property p(o) whose true value varies according to the qualitative criterion [alpha,beta], taup(x,[[alpha,beta]) is presented in this paper. The inner product transformation of qualitative criterion w_[alpha,beta] and the relation between w_[alpha,beta] and artificial neuron is discussed. It is shown that an artificial neuron is just a boundary of a qualitative mapping, and the inner product transformation of qualitative criterion would induce two group of linear transformation which rotation centers is a pair of vertexes of qualitative criterion [betaalpha] respectively
{"title":"Qualitative Mapping and Artificial Neuron","authors":"Jia-li Feng, Guanglin Xu, Zhanqiu Dong, Jingjuan Feng","doi":"10.1109/ICNNB.2005.1614577","DOIUrl":"https://doi.org/10.1109/ICNNB.2005.1614577","url":null,"abstract":"The qualitative mapping (QM) model for judging a property p(o) whose true value varies according to the qualitative criterion [alpha,beta], taup(x,[[alpha,beta]) is presented in this paper. The inner product transformation of qualitative criterion w_[alpha,beta] and the relation between w_[alpha,beta] and artificial neuron is discussed. It is shown that an artificial neuron is just a boundary of a qualitative mapping, and the inner product transformation of qualitative criterion would induce two group of linear transformation which rotation centers is a pair of vertexes of qualitative criterion [betaalpha] respectively","PeriodicalId":145719,"journal":{"name":"2005 International Conference on Neural Networks and Brain","volume":"127 1","pages":"0"},"PeriodicalIF":0.0,"publicationDate":"2005-10-13","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"115091655","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：