作者: Herbivorous 時間: 2025-3-21 21:20
Brent Kawahara,Hector Estrada,Luke S. Leetional invariance can be based, and try to delimit the conditions under which each of them acts. We find out that, surprisingly, some of the most popular neural learning methods, such as weight-decay and input noise addition, exhibit this interesting property.作者: 無法治愈 時間: 2025-3-22 00:24 作者: 愛哭 時間: 2025-3-22 05:00 作者: STALE 時間: 2025-3-22 11:37
https://doi.org/10.1057/978-1-349-93268-9 weight in the output layer is derived as a nonlinear function of the training data moments. The experimental results, using one- and two-dimensional simulated data and different polynomial orders, show that the classification rate of the polynomial densities is very close to the optimum rate.作者: 儀式 時間: 2025-3-22 14:55
Neural Learning Invariant to Network Size Changestional invariance can be based, and try to delimit the conditions under which each of them acts. We find out that, surprisingly, some of the most popular neural learning methods, such as weight-decay and input noise addition, exhibit this interesting property.作者: 尖牙 時間: 2025-3-22 18:34 作者: Etymology 時間: 2025-3-22 21:21
Discriminative Dimensionality Reduction Based on Generalized LVQionality reduction in feature extraction. Experimental results reveal that the training of both a feature transformation matrix and reference vectors by GLVQ is superior to that by principal component analysis in terms of dimensionality reduction.作者: 業(yè)余愛好者 時間: 2025-3-23 04:34 作者: 甜得發(fā)膩 時間: 2025-3-23 08:32
Fast Curvature Matrix-Vector ProductsFisher information matrices with arbitrary vectors, using techniques similar to but even cheaper than the fast Hessian-vector product [.]. The stability of SMD [.,.,.,.], a learning rate adaptation method that uses curvature matrix-vector products, improves when the extended Gauss-Newton matrix is substituted for the Hessian.作者: ARK 時間: 2025-3-23 12:43
Bagging Can Stabilize without Reducing Variance accepted. This paper provides experimental evidence supporting another explanation, based on the stabilization provided by spreading the influence of examples. With this viewpoint, bagging is interpreted as a case-weight perturbation technique, and its behavior can be explained when other arguments fail.作者: 嘮叨 時間: 2025-3-23 14:29
Learning to Learn Using Gradient Descentn large systems feasible by using recurrent neural networks with their attendant learning routines as meta-learning systems. Our system derived complex well performing learning algorithms from scratch. In this paper we also show that our approach performs non-stationary time series prediction.作者: 貨物 時間: 2025-3-23 22:04 作者: 眼界 時間: 2025-3-24 00:25 作者: arboretum 時間: 2025-3-24 05:16
Bruno Crosignani,Benedetto Dainotive function. In this paper, radial basis function networks (RBFN) are employed in predicting the form of objective function, and genetic algorithms (GA) in searching the optimal value of the predicted objective function. The effectiveness of the suggested method will be shown through some numerical examples.作者: bacteria 時間: 2025-3-24 08:44
Exotic Uses of Guided Wave Optics,rete variable. We apply the method to a time series data set, i.e. yeast gene expressions measured with DNA chips, with biological knowledge about the functions of the genes encoded into the discrete variable.作者: indigenous 時間: 2025-3-24 14:05 作者: 寒冷 時間: 2025-3-24 18:23 作者: Kidnap 時間: 2025-3-24 20:13 作者: 神圣在玷污 時間: 2025-3-24 23:56 作者: 啞巴 時間: 2025-3-25 06:59
Generalisation Improvement of Radial Basis Function Networks Based on Qualitative Input Conditioningthe problem is strongly recommended when using learning techniques. The paper aims at conditioning the input information in order to enhance the neural network generalisation by adding qualitative expert information on orders of magnitude. An example of this method applied to some industrial firms is given.作者: 減震 時間: 2025-3-25 08:18
Conference proceedings 2001enna University of Technology, Austria. The conference is organized by the A- trian Research Institute for Arti?cal Intelligence in cooperation with the Pattern Recognition and Image Processing Group and the Center for Computational - telligence at the Vienna University of Technology. The ICANN conf作者: 公式 時間: 2025-3-25 15:35
Brent Kawahara,Hector Estrada,Luke S. LeeFisher information matrices with arbitrary vectors, using techniques similar to but even cheaper than the fast Hessian-vector product [.]. The stability of SMD [.,.,.,.], a learning rate adaptation method that uses curvature matrix-vector products, improves when the extended Gauss-Newton matrix is substituted for the Hessian.作者: Aspirin 時間: 2025-3-25 17:55 作者: 傷心 時間: 2025-3-25 22:07
Optical Waveguide Modulation Techniques,n large systems feasible by using recurrent neural networks with their attendant learning routines as meta-learning systems. Our system derived complex well performing learning algorithms from scratch. In this paper we also show that our approach performs non-stationary time series prediction.作者: 滑動 時間: 2025-3-26 03:38 作者: 縮短 時間: 2025-3-26 07:42
0302-9743 ngs. We would like to thank the European Neural Network Society (ENNS) for their support. We acknowledge the ?nancial support of Austrian Airlines, A- trian Science Foundation (FWF) under the c978-3-540-42486-4978-3-540-44668-2Series ISSN 0302-9743 Series E-ISSN 1611-3349 作者: blister 時間: 2025-3-26 10:13 作者: Foment 時間: 2025-3-26 14:26 作者: 幻想 時間: 2025-3-26 19:52
https://doi.org/10.1057/978-1-349-93268-9he advantage that they are free from such assumptions. They can be, however, sensitive to noise and computationally intensive in high-dimensional spaces. In this paper we re-consider the issue of data partitioning from an information-theoretic viewpoint and show that minimisation of partition entrop作者: GOUGE 時間: 2025-3-26 23:38 作者: voluble 時間: 2025-3-27 03:59 作者: abstemious 時間: 2025-3-27 07:07
The Complementary Brain (Abstract)paradigm that radically departs from the computer metaphor suggesting that brains are organized into independent modules. Evidence is reviewed that brains are organized into parallel processing streams with complementary properties. This perspective clarifies, for example, how parallel processing in作者: MEEK 時間: 2025-3-27 10:45 作者: white-matter 時間: 2025-3-27 15:08 作者: Gentry 時間: 2025-3-27 21:23 作者: 招惹 時間: 2025-3-27 22:12 作者: BALE 時間: 2025-3-28 05:13 作者: 四指套 時間: 2025-3-28 07:47 作者: unstable-angina 時間: 2025-3-28 10:34
Bagging Can Stabilize without Reducing Varianceeld better results than the original predictor. It is thus important to understand the reasons for this success, and also for the occasional failures. Several arguments have been given to explain the effectiveness of bagging, among which the original “bagging reduces variance by averaging” is widely作者: 學術(shù)討論會 時間: 2025-3-28 15:49 作者: 鳥籠 時間: 2025-3-28 20:25
Discriminative Dimensionality Reduction Based on Generalized LVQcognition. GLVQ is a general framework for classifier design based on the minimum classification error criterion, and it is easy to apply it to dimensionality reduction in feature extraction. Experimental results reveal that the training of both a feature transformation matrix and reference vectors 作者: 脆弱么 時間: 2025-3-29 00:21 作者: 叫喊 時間: 2025-3-29 06:30
Clustering Gene Expression Data by Mutual Information with Gene Function space and become local there, while within-cluster differences between the associated, implicitly estimated conditional distributions of the discrete variable are minimized. The discrete variable can be seen as an indicator of relevance or importance guiding the clustering. Minimization of the Kull作者: Itinerant 時間: 2025-3-29 08:46 作者: endocardium 時間: 2025-3-29 13:53 作者: eczema 時間: 2025-3-29 18:06 作者: 匍匐 時間: 2025-3-29 22:42 作者: 很是迷惑 時間: 2025-3-30 01:05 作者: Insul島 時間: 2025-3-30 05:42
Approximation of Bayesian Discriminant Function by Neural Networks in Terms of Kullback-Leibler Infoork, having rather a small number of hidden layer units, can approximate the Bayesian discriminant function for the two category classification if the log ratio of the a posteriori probability is a polynomial. The accuracy of approximation is measured by the Kullback-Leibler information. An extensio作者: 吞沒 時間: 2025-3-30 08:23
Neural Networks for Adaptive Processing of Structured Datahe ability to recognize and classify these patterns is fundamental for several applications that use, generate or manipulate structures. In this paper I review some of the concepts underpinning Recursive Neural Networks, i.e. neural network models able to deal with data represented as directed acyclic graphs.作者: 鄙視讀作 時間: 2025-3-30 12:57 作者: Outshine 時間: 2025-3-30 19:32
Architecture Selection in NLDA Networksork we study the architecture selection problem for NLDA networks. We shall derive asymptotic distribution results for NLDA weights, from which Wald like tests can be derived. We also discuss how to use them to make decisions on unit relevance based on the acceptance or rejection of a certain null hypothesis.作者: 積極詞匯 時間: 2025-3-30 22:15
Approximation of Bayesian Discriminant Function by Neural Networks in Terms of Kullback-Leibler Infoork, having rather a small number of hidden layer units, can approximate the Bayesian discriminant function for the two category classification if the log ratio of the a posteriori probability is a polynomial. The accuracy of approximation is measured by the Kullback-Leibler information. An extension to the multi-category case is also discussed.作者: Axon895 時間: 2025-3-31 04:26
https://doi.org/10.1007/3-540-44668-0Artificial neural networks; Beowulf; Controller Area Network (CAN); Online; Segment; artificial intellige作者: poliosis 時間: 2025-3-31 07:38
978-3-540-42486-4Springer-Verlag Berlin Heidelberg 2001作者: Condescending 時間: 2025-3-31 11:24 作者: photopsia 時間: 2025-3-31 13:27 作者: 厚臉皮 時間: 2025-3-31 20:40 作者: Monocle 時間: 2025-3-31 23:54 作者: molest 時間: 2025-4-1 03:31 作者: jettison 時間: 2025-4-1 09:18 作者: 顧客 時間: 2025-4-1 12:15
Lecture Notes in Computer Sciencehttp://image.papertrans.cn/b/image/162701.jpg作者: kyphoplasty 時間: 2025-4-1 14:34 作者: BRAND 時間: 2025-4-1 21:26
Modification of Cellulose Acetate Filmshe ability to recognize and classify these patterns is fundamental for several applications that use, generate or manipulate structures. In this paper I review some of the concepts underpinning Recursive Neural Networks, i.e. neural network models able to deal with data represented as directed acycl作者: 得意牛 時間: 2025-4-2 02:43 作者: prosthesis 時間: 2025-4-2 05:47
Brent Kawahara,Hector Estrada,Luke S. Leermation. We extend it from nonlinear least squares to all differentiable objectives such that positive semi-definiteness is maintained for the standard loss functions in neural network regression and classification. We give efficient algorithms for computing the product of extended Gauss-Newton and
