[下载]一本介绍神经网络的书

3202

收藏 2008-01-17

<An Introduction to Neural Networks> CONTENTS Preface I FUNDAMENTALS 1 Introduction 2 Fundamentals 2.1 A framework for distributed representation 2.1.1 Processing units 2.1.2 Connections between units 2.1.3 Activation and output rules 2.2 Network topologies 2.3 Training of arti_cial neural networks 2.3.1 Paradigms of learning 2.3.2 Modifying patterns of connectivity 2.4 Notation and terminology 2.4.1 Notation 2.4.2 Terminology II THEORY 3 Perceptron and Adaline 3.1 Networks with threshold activation functions 3.2 Perceptron learning rule and convergence theorem 3.2.1 Example of the Perceptron learning rule 3.2.2 Convergence theorem 3.2.3 The original Perceptron 3.3 The adaptive linear element (Adaline) 3.4 Networks with linear activation functions the delta rule 3.5 Exclusive-OR problem 3.6 Multi-layer perceptrons can do everything 3.7 Conclusions 4 Back-Propagation 4.1 Multi-layer feed-forward networks 4.2 The generalised delta rule 4.2.1 Understanding back-propagation 4.3 Working with back-propagation 4.4 An example 4.5 Other activation functions 4.6 De_ciencies of back-propagation 4.7 Advanced algorithms 4.8 How good are multi-layer feed-forward networks? 4.8.1 The e_ect of the number of learning samples 4.8.2 The e_ect of the number of hidden units 4.9 Applications 5 Recurrent Networks 5.1 The generalised delta-rule in recurrent networks 5.1.1 The Jordan network 5.1.2 The Elman network 5.1.3 Back-propagation in fully recurrent networks 5.2 The Hop_eld network 5.2.1 Description 5.2.2 Hop_eld network as associative memory 5.2.3 Neurons with graded response 5.3 Boltzmann machines 6 Self-Organising Networks 6.1 Competitive learning 6.1.1 Clustering 6.1.2 Vector quantisation 6.2 Kohonen network 6.3 Principal component networks 6.3.1 Introduction 6.3.2 Normalised Hebbian rule 6.3.3 Principal component extractor 6.3.4 More eigenvectors 6.4 Adaptive resonance theory 6.4.1 Background Adaptive resonance theory 6.4.2 ART1 The simpli_ed neural network model 6.4.3 ART1 The original model 7 Reinforcement learning 7.1 The critic 7.2 The controller network 7.3 Barto's approach the ASE-ACE combination 7.3.1 Associative search 7.3.2 Adaptive critic 7.3.3 The cart-pole system 7.4 Reinforcement learning versus optimal control III APPLICATIONS 8 Robot Control 8.1 End-e_ector positioning 8.1.1 Camera{robot coordination is function approximation 8.2 Robot arm dynamics 8.3 Mobile robots 8.3.1 Model based navigation 8.3.2 Sensor based control 9 Vision 9.1 Introduction 9.2 Feed-forward types of networks 9.3 Self-organising networks for image compression 9.3.1 Back-propagation 9.3.2 Linear networks 9.3.3 Principal components as features 9.4 The cognitron and neocognitron 9.4.1 Description of the cells 9.4.2 Structure of the cognitron 9.4.3 Simulation results 9.5 Relaxation types of networks 9.5.1 Depth from stereo 9.5.2 Image restoration and image segmentation 9.5.3 Silicon retina IV IMPLEMENTATIONS 10 General Purpose Hardware 10.1 The Connection Machine 10.1.1 Architecture 10.1.2 Applicability to neural networks 10.2 Systolic arrays 11 Dedicated Neuro-Hardware 11.1 General issues 11.1.1 Connectivity constraints 11.1.2 Analogue vs. digital 11.1.3 Optics 11.1.4 Learning vs. non-learning 11.2 Implementation examples 11.2.1 Carver Mead's silicon retina 11.2.2 LEP's LNeuro chip References Index http://down8931.pinggu.org/UploadFile_20082009/2008-10/200810120505383402.pdf

[此贴子已经被作者于2008-10-12 1:04:31编辑过]