A novel structure learning algorithm for fuzzy neural networks (SLNN) is presented in this paper. The neurons of SLNN are created and adapted as online learning proceeds. The learning rule of SLNN is based on Hebbian ...
详细信息
A novel structure learning algorithm for fuzzy neural networks (SLNN) is presented in this paper. The neurons of SLNN are created and adapted as online learning proceeds. The learning rule of SLNN is based on Hebbian learning and a kernel winner-take-all algorithm - KWTA. KWTA not only can let SLNN be able to learn from new data but also can prevent losing the knowledge which has been learned earlier. To obtain a concise fuzzy rule, a pruning algorithm is adopted in SLNN which doesn't disobey the basic design philosophy of fuzzy system. simulations are performed on the primary benchmark: circle-in-the-square. Comparison with ARTMAP and BP neural network indicates that better performance is achieved
The diameter protocol is recommended by IETF as AAA (authentication, authorization and accounting) protocol criterion for the next generation network. Because the IPv6 protocol will be widely applied in the intending ...
详细信息
The diameter protocol is recommended by IETF as AAA (authentication, authorization and accounting) protocol criterion for the next generation network. Because the IPv6 protocol will be widely applied in the intending all-IP network, mobile IPv6 application based on diameter protocol will play more important role in authentication, authorization and accounting. In this paper, the implementation of mobile node's authentication and authorization is presented with PANA (protocol for carrying authentication for network access) protocol. It is based on diameter protocol for the application expansion of mobile IPv6, which provides the supports to the basic AAA process of mobile IPv6 nodes and dynamic home agent distribution in the visited network and the secret key distribution. Finally, the correctness of this application expansion is testified with developing the design of protocol based on opendiameter
In this paper a new technique for computing and ray tracing point based geometry is presented. It uses a novel point primitive that is called "Spherical Patch Point" (SPP) to approximate the vicinity of a su...
详细信息
In this paper a new technique for computing and ray tracing point based geometry is presented. It uses a novel point primitive that is called "Spherical Patch Point" (SPP) to approximate the vicinity of a surface point. Due to property of curvature, SPP can achieve similar visual quality compared with previous methods with much fewer primitives for the cost of a few additional bytes per point and thus makes a significant reduction in rendering time. During pre-process,important attributes are added to each SPP for the purpose of ray tracing. During rendering, an intersection algorithm different from previous ones has been demonstrated to get satisfied results. The proposed technique makes it possible to render high quality ray traced images with global illumination using SPPs. It offers a higher ray tracing speed in comparison with previous methods.
In this paper a new point-based rendering method for ray tracing is presented. An oriented spherical patch that passes a surface point is used to approximate the vicinity of that *** this paper the spherical patch tog...
详细信息
In this paper a new point-based rendering method for ray tracing is presented. An oriented spherical patch that passes a surface point is used to approximate the vicinity of that *** this paper the spherical patch together with the surface point is called "Spherical Patch Point" (SPP). Due to property of curvature, SPP can achieve similar visual quality compared with previous methods with much fewer points. This paper defines new point attributes for the purpose of efficiently locating the intersection between incoming ray and ***, an algorithm of intersecting a ray with point geometry is proposed. The algorithm can achieve a higher rendering speed in comparison with previous methods. The presented technique deals well with shadow, reflection and refraction.
Identification of transcription factor binding sites (TFBS) from the upstream region of genes remains a highly important and unsolved problem particularly in higher eukaryotic genomes. In this paper, we propose a new ...
详细信息
Identification of transcription factor binding sites (TFBS) from the upstream region of genes remains a highly important and unsolved problem particularly in higher eukaryotic genomes. In this paper, we propose a new approach to predict TFBS. This approach uses position weight matrix (PWM) to represent binding sites and uses genetic algorithm (GA) to search the best matrix. A new coding method so called multiple-variable coding is proposed in GA. We apply it on two transcription factors rebl and mgl. The result shows that this approach can find most of the known sites, which indicates that this method is very effective
Cancer classification and identification are major areas in medical research. DNA microarrays could provide useful information for cancer classification at the gene expression level. The number of genes in a microarra...
详细信息
Cancer classification and identification are major areas in medical research. DNA microarrays could provide useful information for cancer classification at the gene expression level. The number of genes in a microarray is always several thousands while the number of training samples always several dozens. In such case most of the machine learning models suffer from the overfitting and it is necessary to select a handful of most informative genes. An adaptive and iterative gene selection algorithm based on least squares support vector machines is proposed in this paper. The algorithm adopts sequential forward selection search scheme. The number of selected genes can be determined adaptively. The total number of genes processed by the proposed algorithm is smaller than that processed by other algorithms using support vector machines. Results of numerical experiments show that the proposed algorithm trains fast and achieves comparable performance on two well-known benchmark problems.
An approach to design optimal auto-disturbance-rejection controller was presented for a class of observable high-order nonlinear plant. The controller parameters were optimized under the universal frame of genetic alg...
详细信息
An approach to design optimal auto-disturbance-rejection controller was presented for a class of observable high-order nonlinear plant. The controller parameters were optimized under the universal frame of genetic algorithms, and the desired dynamical equations were obtained according to the specified performance indices at the same time. A high-order plant was simulated using the method. Results were compared with the PID and exact feedback linearization control system. The results demonstrate that the method performs powerful functions for improving the dynamic characteristics. simulation results also show that the method has good robustness against disturbances, noises and modeling uncertainties
To improve artificial intelligence (AI) of computer games is a hard problem. Qualitative spatial reasoning is utilized to solve this problem. In qualitative spatial reasoning various aspects of qualitative spatial rea...
详细信息
To improve artificial intelligence (AI) of computer games is a hard problem. Qualitative spatial reasoning is utilized to solve this problem. In qualitative spatial reasoning various aspects of qualitative spatial reasoning such as topology, direction, size and distance have been widely investigated in pervious literatures. Although, the combination works of two or more spatial aspects are more useful than the single one in computer game and other applications, most combining problems have not been discussed before. So a unified model for qualitative topology and distance information is proposed. Finally this method is applied to strategy computer games.
A dynamic growing neural network (DGNN) for supervised learning of pattern recognition or unsupervised learning of clustering is presented. The main ideas included in DGNN are growing, resonance, and post-prune. DGNN ...
详细信息
A dynamic growing neural network (DGNN) for supervised learning of pattern recognition or unsupervised learning of clustering is presented. The main ideas included in DGNN are growing, resonance, and post-prune. DGNN is called dynamic growing because it is based on the Hebbian learning rule and adds new neurons under certain conditions. When DGNN performs supervised learning, resonance will happen if the winner can't match the training example; this rule combines the ART/ARTMAP neural network and WTA learning rule. When DGNN performs unsupervised learning, post-prune is carried out to prevent over fitting the training data just like decision tree learning. DGNN's prune rule is based on the distance threshold. DGNN has some advantages: learning not only is stable because it grows under certain conditions; but also it is faster than back-propagation rules and favorable learned predictive accuracy in small, noisy, online or offline data sets. Three classes of simulations are performed on the primary benchmarks: circle-in-the-square and two-spirals-apart benchmarks are used to check DGNN's supervised learning and compare it with ARTMAP and BP neural networks; DGNN's unsupervised learning ability is checked on UCI Machine Learning Archive's Synthetic Control Chart Time Series data set
暂无评论