Neural network experiments on personal computers and workstations
著者
書誌事項
Neural network experiments on personal computers and workstations
MIT Press, c1991
大学図書館所蔵 全22件
  青森
  岩手
  宮城
  秋田
  山形
  福島
  茨城
  栃木
  群馬
  埼玉
  千葉
  東京
  神奈川
  新潟
  富山
  石川
  福井
  山梨
  長野
  岐阜
  静岡
  愛知
  三重
  滋賀
  京都
  大阪
  兵庫
  奈良
  和歌山
  鳥取
  島根
  岡山
  広島
  山口
  徳島
  香川
  愛媛
  高知
  福岡
  佐賀
  長崎
  熊本
  大分
  宮崎
  鹿児島
  沖縄
  韓国
  中国
  タイ
  イギリス
  ドイツ
  スイス
  フランス
  ベルギー
  オランダ
  スウェーデン
  ノルウェー
  アメリカ
注記
"A Bradford book."
Includes bibliographical references and index
内容説明・目次
内容説明
Most neural network programs for personal computers simply control a set of fixed, canned network-layer algorithms with pulldown menus. This tutorial offers hands-on neural network experiments with a different approach. A simple matrix language lets users create their own neural networks and combine networks, and this software also permits combined simulation of neural networks together with other dynamic systems such as robots or physiological models. The enclosed student version of DESIRE/NEUNET differs from the full system only in the size of its data area and includes a screen editor, compiler, colour graphics, help screens, and ready-to-run examples. Users can also add their own help screens and interactive menus. The book provides an introduction to neural networks and simulation, a tutorial on the software, and many complete programs including several backpropagation schemes, creeping random search, competitive learning with and without adaptive-resonance function and "conscience" counterpropagation, nonlinear Grossberg-type neurons, Hopfield-type and bidirectional associative memories, predictors, function learning, biological clocks, system identification and more.
In addition, the book introduces a simple, integrated environment for programming, displays and report preparation. Even differential equations are entered in ordinary mathematical notation. Users need not learn C or LISP to program nonlinear neuron models. To permit truly interactive experiments, the extra-fast compilation is unnoticeable, and simulations execute faster than PC FORTRAN. The nearly 90 illustrations include block diagrams, computer programs, and simulation-output graphs.
目次
- Part 1 Introduction to neural networks and simulation: neural networks and neuron models
- neural networks - decision making and abstraction
- computer representation of neural networks
- true interactive simulation requires direct program execution
- the user-friendly simulation environment. Part 2 A simple language for neural network models: neuron-layer arrays and simulation runs
- vector assignments in neuron-layer models
- simple difference equations specify neuron dynamics
- operations on connection-weight matrices
- DOT products, vector norms, error measures, and sums
- vector compiler and model generality
- scalar expressions in dynamic program segments
- simulation output - desire/neunet graphics and listings. Part 3 Feedforward networks for supervised training: improving on perceptron learning
- simulation of backpropagation networks
- creeping random search and higher-order connections. Part 4 Competitive learning, adaptive-resonance emulation, and statistical learning: unsupervised pattern classification
- the need for adaptive-resonance type schemes
- supervised learning - counterpropagation networks
- recall with random inputs
- statistical learning. Part 5 More advanced neural network modelling: special modelling techniques
- the vector shift operator and simple array convolutions
- contrast enhancement, pattern normalization, and biological models. Part 6 Neural networks with feedback - associative memories, stability and learning: some classical associative memories
- bidirectional associative memories
- network stability and pattern associator capacity. Part 7 More general networks, and combined simulation of neural networks and other dynamic systems: function-mapping neural networks with random inputs and non-least-squares optimization
- sequential neural networks and more general feedback
- models involving differential equations
- simulations combining neural networks and separate differential-equation systems. Appendices: Desire / Neunet programs and files - hardware requirements and software installation, Desire / Neunet commands and user programs, program entry, the two desire screen editors, file manipulation, combining and chaining programme segments
- interpreted experiment-protocol programs - a simple interpreter language, special features for interactive experiments, Desire / Neunet library functions.
「Nielsen BookData」 より