River formation dynamics (RFD) is an heuristic optimization algorithm based on copying how water forms rivers by eroding the ground and depositing sediments. After drops transform the landscape by increasing/decreasing the altitude of places, solutions are given in the form of paths of decreasing altitudes. Decreasing gradients are constructed, and these gradients are followed by subsequent drops to compose new gradients and reinforce the best ones. We apply this method to solve dynamic TSP. We show that the gradient orientation of RFD makes it specially suitable for solving this problem, and we compare our results with those given by ant colony optimization (ACO).
{"title":"Solving Dynamic TSP by Using River Formation Dynamics","authors":"Pablo Rabanal, Ismael Rodríguez, F. Rubio","doi":"10.1109/ICNC.2008.760","DOIUrl":"https://doi.org/10.1109/ICNC.2008.760","url":null,"abstract":"River formation dynamics (RFD) is an heuristic optimization algorithm based on copying how water forms rivers by eroding the ground and depositing sediments. After drops transform the landscape by increasing/decreasing the altitude of places, solutions are given in the form of paths of decreasing altitudes. Decreasing gradients are constructed, and these gradients are followed by subsequent drops to compose new gradients and reinforce the best ones. We apply this method to solve dynamic TSP. We show that the gradient orientation of RFD makes it specially suitable for solving this problem, and we compare our results with those given by ant colony optimization (ACO).","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"42 1","pages":"246-250"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89125397","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 is concerned with stochastic robust stability of a class of stochastic neural networks with time varying delays and parameter uncertainties. The parameter uncertainties are time-varying and norm-bounded. Based on Lyapunov-Krasovskii functional and stochastic analysis approaches, new stability criteria are presented in terms of linear matrix inequalities (LMIs) to guarantee the delayed neural network to be robustly stochastically asymptotically stable in the mean square for all admissible uncertainties. Numerical examples are given to demonstrate the usefulness of the proposed robust stability criteria.
{"title":"Robust Stability Analysis of Uncertain Stochastic Neural Networks with Time-Varying Delays","authors":"W. Feng, Wei Zhang, Haixia Wu","doi":"10.1109/ICNC.2008.568","DOIUrl":"https://doi.org/10.1109/ICNC.2008.568","url":null,"abstract":"This paper is concerned with stochastic robust stability of a class of stochastic neural networks with time varying delays and parameter uncertainties. The parameter uncertainties are time-varying and norm-bounded. Based on Lyapunov-Krasovskii functional and stochastic analysis approaches, new stability criteria are presented in terms of linear matrix inequalities (LMIs) to guarantee the delayed neural network to be robustly stochastically asymptotically stable in the mean square for all admissible uncertainties. Numerical examples are given to demonstrate the usefulness of the proposed robust stability criteria.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"75 2 1","pages":"522-526"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"89213606","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 makes a discussion of consistent subsets (CS) selection criteria for hyper surface Classification (HSC) and SVM algorithms. The consistent subsets play an important role in the data selection. Firstly, the paper proposes that minimal consistent subset for a disjoint cover set (MCSC) plays an important role in the data selection for HSC. The MCSC can be applied to select a representative subset from the original sample set for HSC. MCSC has the same classification model with the entire sample set and can totally reflect its classification ability. Secondly, the number of MCSC is calculated. Thirdly, by comparing the performance of HSC and SVM on corresponding CS, we argue that it is not reasonable that using the same train data set to train different classifiers and then testing the classifiers by the same test data set for different algorithms. The experiments show that algorithms can respectively select the proper data set for training, which ensures good performance and generalization ability. MCSC is the best selection for HSC, and support vector set is the effective selection for SVM.
{"title":"The Data Selection Criteria for HSC and SVM Algorithms","authors":"Qing He, Fuzhen Zhuang, Zhongzhi Shi","doi":"10.1109/ICNC.2008.334","DOIUrl":"https://doi.org/10.1109/ICNC.2008.334","url":null,"abstract":"This paper makes a discussion of consistent subsets (CS) selection criteria for hyper surface Classification (HSC) and SVM algorithms. The consistent subsets play an important role in the data selection. Firstly, the paper proposes that minimal consistent subset for a disjoint cover set (MCSC) plays an important role in the data selection for HSC. The MCSC can be applied to select a representative subset from the original sample set for HSC. MCSC has the same classification model with the entire sample set and can totally reflect its classification ability. Secondly, the number of MCSC is calculated. Thirdly, by comparing the performance of HSC and SVM on corresponding CS, we argue that it is not reasonable that using the same train data set to train different classifiers and then testing the classifiers by the same test data set for different algorithms. The experiments show that algorithms can respectively select the proper data set for training, which ensures good performance and generalization ability. MCSC is the best selection for HSC, and support vector set is the effective selection for SVM.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"30 1","pages":"384-388"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"90220173","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 convention method of testing breakwaters in flume by physical model was replaced by purely numerical model. A Bretschneider/Mitsuyasu spectrum wave was generated in numerical flume, and the design of breakwater is in accordance with Code of Hydrology for Sea Harbour, China ("the Code"). Modified Boussinesq type wave equations were solved including a wave reflection term. By purely simulation method, we have a reflection coefficient of 0.45, and in comparison with the Code, it is concluded numerical models can be used in the pre-design of a breakwater, and safety coefficient is used in the Code.
{"title":"Numerical Simulation of Nonliner Wave Propagating in Flume","authors":"Su-Xiang Zhang, Xi Li","doi":"10.1109/ICNC.2008.854","DOIUrl":"https://doi.org/10.1109/ICNC.2008.854","url":null,"abstract":"The convention method of testing breakwaters in flume by physical model was replaced by purely numerical model. A Bretschneider/Mitsuyasu spectrum wave was generated in numerical flume, and the design of breakwater is in accordance with Code of Hydrology for Sea Harbour, China (\"the Code\"). Modified Boussinesq type wave equations were solved including a wave reflection term. By purely simulation method, we have a reflection coefficient of 0.45, and in comparison with the Code, it is concluded numerical models can be used in the pre-design of a breakwater, and safety coefficient is used in the Code.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"25 1","pages":"649-653"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79077371","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}
Qu Li, Weihong Wang, Xing Qi, Bo Chen, Jianhong Li
Gene expression programming (GEP) is a kind of heuristic method based on evolutionary computation theory. Basic GEP method has been proved to be powerful in symbolic regression and other data mining as well as machine learning tasks. However, GEP's potential for neural network learning has not been well studied. In this paper, we prove that GEP neural network (GEPNN) is not able to solve high order regression problems. Based on our proof, we propose an extended method for evolving neural network with GEP. The extended GEPNN is used in various kinds of function finding problems. Results on multiple leaning methods show the effectiveness of our method.
{"title":"Function Finding Using Gene Expression Programming Based Neural Network","authors":"Qu Li, Weihong Wang, Xing Qi, Bo Chen, Jianhong Li","doi":"10.1109/ICNC.2008.688","DOIUrl":"https://doi.org/10.1109/ICNC.2008.688","url":null,"abstract":"Gene expression programming (GEP) is a kind of heuristic method based on evolutionary computation theory. Basic GEP method has been proved to be powerful in symbolic regression and other data mining as well as machine learning tasks. However, GEP's potential for neural network learning has not been well studied. In this paper, we prove that GEP neural network (GEPNN) is not able to solve high order regression problems. Based on our proof, we propose an extended method for evolving neural network with GEP. The extended GEPNN is used in various kinds of function finding problems. Results on multiple leaning methods show the effectiveness of our method.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"61 1","pages":"195-198"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79201227","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}
Dong-Chul Park, Nhon Huu Tran, Dong-Min Woo, Yunsik Lee
A kernel-based centroid neural network with spatial constraints (K-CNN-S) is proposed and presented in this paper. The proposed K-CNN-S is based on the centroid neural network (CNN) and also exploits advantages of the kernel method for mapping input data into a higher dimensional feature space. Furthermore, The K-CNN-S adopts the spatial constraints to reduce noise in images. The magnetic resonance image (MRI) segmentation is performed to illustrate the application of the proposed K-CNN-S algorithm. Experiments and results on MRI data from Internet brain segmentation repository (IBSR) demonstrate that image segmentation scheme based on the proposed K-CNN-S outperforms conventional algorithms including fuzzy c-means (FCM), kernel-based fuzzy c-mean (K-FCM), and kernel-based fuzzy c-mean with spatial constraints (K-FCM-S).
{"title":"Kernel-Based Centroid Neural Network with Spatial Constraints for Image Segmentation","authors":"Dong-Chul Park, Nhon Huu Tran, Dong-Min Woo, Yunsik Lee","doi":"10.1109/ICNC.2008.635","DOIUrl":"https://doi.org/10.1109/ICNC.2008.635","url":null,"abstract":"A kernel-based centroid neural network with spatial constraints (K-CNN-S) is proposed and presented in this paper. The proposed K-CNN-S is based on the centroid neural network (CNN) and also exploits advantages of the kernel method for mapping input data into a higher dimensional feature space. Furthermore, The K-CNN-S adopts the spatial constraints to reduce noise in images. The magnetic resonance image (MRI) segmentation is performed to illustrate the application of the proposed K-CNN-S algorithm. Experiments and results on MRI data from Internet brain segmentation repository (IBSR) demonstrate that image segmentation scheme based on the proposed K-CNN-S outperforms conventional algorithms including fuzzy c-means (FCM), kernel-based fuzzy c-mean (K-FCM), and kernel-based fuzzy c-mean with spatial constraints (K-FCM-S).","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"15 3 1","pages":"236-240"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"79273336","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 achievements of computer graphics have supported the developments of augmented reality (AR), a technique of an automatic connection of additional information generated from a computer model to reality, which is offered to enhance or augment the cognition to the real world of a user. Especially due to the developments in hardware in the recent years not only visualization but also interaction with these environments has become possible. Outdoor AR system integrates the presentation of three-dimensional objects out of a database on site in a Geographical information system (GIS). It not only expands a GIS to the third dimension, also reacts to the current position and viewing field of the user. From the perspective of the theory of spatial cognition and visualization, the paper puts forward there search contents and basic structure of spatial data AR visualization by analyzing the research status and related research works, comparing there exists frameworks. Furthermore it sums up the key technologies in detail and proposes the improved approaches. By a case study, it verifies the procedure of AR visualization of geospatial data. Lastly, it discusses the significant applications and development prospects of AR visualization of geospatial data.
{"title":"Cognition-Based Augmented Reality Visualization of the Geospatial Data","authors":"Xueling Wu, Qingyun Du, F. Ren","doi":"10.1109/ICNC.2008.814","DOIUrl":"https://doi.org/10.1109/ICNC.2008.814","url":null,"abstract":"The achievements of computer graphics have supported the developments of augmented reality (AR), a technique of an automatic connection of additional information generated from a computer model to reality, which is offered to enhance or augment the cognition to the real world of a user. Especially due to the developments in hardware in the recent years not only visualization but also interaction with these environments has become possible. Outdoor AR system integrates the presentation of three-dimensional objects out of a database on site in a Geographical information system (GIS). It not only expands a GIS to the third dimension, also reacts to the current position and viewing field of the user. From the perspective of the theory of spatial cognition and visualization, the paper puts forward there search contents and basic structure of spatial data AR visualization by analyzing the research status and related research works, comparing there exists frameworks. Furthermore it sums up the key technologies in detail and proposes the improved approaches. By a case study, it verifies the procedure of AR visualization of geospatial data. Lastly, it discusses the significant applications and development prospects of AR visualization of geospatial data.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"1 1","pages":"138-142"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83023293","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}
Hong Zhu, Minhua Wu, Guixia Guan, Yong Guan, Weizhen Sun
Dead reckoning system (DR) and Global Positioning System (GPS), which consist of integrated navigation system, are two important positioning methods in the intelligent vehicle navigation. The information from the different sensors of vehicle GPS and DR integrated navigation system needs to be fused in order to implement the optimal evaluation of global states, because of the different measurements and their noise characteristics. The federal Kalman filter is designed to fuse GPS and DR information. Two local filters process GPS and DR data respectively, and the main filter is responsible for data fusion and reset to the local filters. The information fusion based on federal filter solves some key problems such as system unavailability, big accumulative errors with GPS or DR alone, and it makes the system's global evaluation optimal. The simulation results show that the positioning accuracy and the credibility of the vehicle integrated navigation are much higher than that when GPS or DR is used alone.
{"title":"Vehicle Multi-sensor Information Optimization Based on Federal Fusion Valuation","authors":"Hong Zhu, Minhua Wu, Guixia Guan, Yong Guan, Weizhen Sun","doi":"10.1109/ICNC.2008.366","DOIUrl":"https://doi.org/10.1109/ICNC.2008.366","url":null,"abstract":"Dead reckoning system (DR) and Global Positioning System (GPS), which consist of integrated navigation system, are two important positioning methods in the intelligent vehicle navigation. The information from the different sensors of vehicle GPS and DR integrated navigation system needs to be fused in order to implement the optimal evaluation of global states, because of the different measurements and their noise characteristics. The federal Kalman filter is designed to fuse GPS and DR information. Two local filters process GPS and DR data respectively, and the main filter is responsible for data fusion and reset to the local filters. The information fusion based on federal filter solves some key problems such as system unavailability, big accumulative errors with GPS or DR alone, and it makes the system's global evaluation optimal. The simulation results show that the positioning accuracy and the credibility of the vehicle integrated navigation are much higher than that when GPS or DR is used alone.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"22 1","pages":"581-585"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"83285970","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}
Yong-Zhi Li, Jing-yu Yang, Songsong Wu, Fen-Xiang Liu
This paper proposes a two-phase algorithm of image projection discriminant analysis. The new discriminant method is composed of feature extraction by on maximum margin criterion (MMC) and Fisher discriminant analysis (FDA). The algorithm includes two stages: firstly, the feature extraction based on maximum margin criterion (MMC) is employed to condense the dimension of image matrix; Then Fisher discriminant analysis (FDA) is applied to reduce dimension of condensed image matrices. This novel method based on image matrix is called 2DMMCplu2DFDA in the paper. Different from the previous linear discriminant analysis method for face recognition where FDA or PCA is based on image vector, 2DMMCplus2DFDA is to exploit image matrices to directly construct the between-class scatter matrix, within-class scatter matrix and total-class scatter matrix. The experimental results on ORL face databases indicate that the proposed method is more efficient and stable than 2DPCA, 2DMMC and 2DPCAplus2DFDA with higher recognition rate.
{"title":"A New Two-Phase Method of Face Recognition Based on Image Matrix","authors":"Yong-Zhi Li, Jing-yu Yang, Songsong Wu, Fen-Xiang Liu","doi":"10.1109/ICNC.2008.562","DOIUrl":"https://doi.org/10.1109/ICNC.2008.562","url":null,"abstract":"This paper proposes a two-phase algorithm of image projection discriminant analysis. The new discriminant method is composed of feature extraction by on maximum margin criterion (MMC) and Fisher discriminant analysis (FDA). The algorithm includes two stages: firstly, the feature extraction based on maximum margin criterion (MMC) is employed to condense the dimension of image matrix; Then Fisher discriminant analysis (FDA) is applied to reduce dimension of condensed image matrices. This novel method based on image matrix is called 2DMMCplu2DFDA in the paper. Different from the previous linear discriminant analysis method for face recognition where FDA or PCA is based on image vector, 2DMMCplus2DFDA is to exploit image matrices to directly construct the between-class scatter matrix, within-class scatter matrix and total-class scatter matrix. The experimental results on ORL face databases indicate that the proposed method is more efficient and stable than 2DPCA, 2DMMC and 2DPCAplus2DFDA with higher recognition rate.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"22 1","pages":"48-52"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"81207224","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}
Yinxue Zhang, Zhenhong Jia, Haijun Jiang, Zijian Liu
A new method for image restoration based on robust error function and BP neural network optimized with particle swarm optimization (PSO) is proposed in this paper. In this technique, BP neural network uses a robust error function as its error function, and then the neural network optimized with PSO. This method can minimize an evaluation function established based on an observed image. The proposed method takes into consideration point spread function (PSF) blurring as well as an additive random noise and obtains restoration image with more preserved image details. Experimental results demonstrate that the proposed new method can have a very high quality both in the visual qualitative performance and the quantitative performance than the traditional algorithms.
{"title":"Image Restoration Based on Robust Error Function and Particle Swarm Optimization-BP Neural Network","authors":"Yinxue Zhang, Zhenhong Jia, Haijun Jiang, Zijian Liu","doi":"10.1109/ICNC.2008.140","DOIUrl":"https://doi.org/10.1109/ICNC.2008.140","url":null,"abstract":"A new method for image restoration based on robust error function and BP neural network optimized with particle swarm optimization (PSO) is proposed in this paper. In this technique, BP neural network uses a robust error function as its error function, and then the neural network optimized with PSO. This method can minimize an evaluation function established based on an observed image. The proposed method takes into consideration point spread function (PSF) blurring as well as an additive random noise and obtains restoration image with more preserved image details. Experimental results demonstrate that the proposed new method can have a very high quality both in the visual qualitative performance and the quantitative performance than the traditional algorithms.","PeriodicalId":6404,"journal":{"name":"2008 Fourth International Conference on Natural Computation","volume":"30 1","pages":"640-644"},"PeriodicalIF":0.0,"publicationDate":"2008-10-18","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"84683849","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：