Neural networks and fuzzy-logic control on personal computers and workstations
Author(s)
Bibliographic Information
Neural networks and fuzzy-logic control on personal computers and workstations
MIT Press, c1995
Available at / 24 libraries
-
No Libraries matched.
- Remove all filters.
Note
"A Bradford book"
Computer disc (the version of DESIRE/NEUNET included here is for PCs, viz, 286/287, 386/387, 486Dx, Pentium, P6, SX with math coprocessor
Rev. ed. of: Neural network experiments on personal computers and workstations, 1991
Includes bibliographical references and index
Description and Table of Contents
Description
Most neural-network programmes for personal computers and engineering workstations simply control a fixed set of canned network-layer algorithms with pulldown menus. This hands-on tutorial demonstrates both neural networks and fuzzy-logic control with an alternative approach. A natural, computer-readable notation for matrix operations and differential equations lets users create their own neural networks and fuzzy-logic controllers on the screen; real simulation experiments then execute immediately. "Neural Networks and Fuzzy-Logic Control" introduces a simple integrated environment for programming displays and report generation. It includes the only currently available software that permits combined simulation of multiple neural networks, fuzzy logic controllers, and dynamic systems such as robots or physiological models. The enclosed educational version of DESIRE/NEUNET differs from the full system mainly in the size of its data area and includes two screen editors, compiler, colour graphics, and many ready-to-run examples. The software lets users or instructors add their own help screens and interactive menus.
Differential equations in scalar and/or matrix form are entered in ordinary mathematical notation. Users can programme new neural networks or fuzzy logic applications without learning C or LISP. For truly interactive experimentation, the extra-fast compilation in unnoticeable, and simulation speed still compares well with that of special accessory processors. The 123 figures include block diagrams, simulation-output graphs produced on personal computers and engineering workstations and many complete computer programmes. The version of DESIRE/NEUNET indluded here is for PCs, viz. 286/287, 386/387, 486DX, Pentium, P6, SX with math coprocessor.
Table of Contents
- 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 programme execution
- the user-friendly simulation environment. Part 2 A readable language of neural network models: interpreted experiment protocol and compiled simulation runs
- dynamic programme segments, neuron-layer models and vector operations
- simple difference equations model neuron dynamics
- operations on connection-weight matrices
- dot products, vector norms, error measures and sums
- more programme features. Part 3 Pattern recognition with neural networks: pattern classification without learning
- neural networks for pattern classification and associative memory
- adaptive neural networks and learning. Part 4 Simple feedforward networks for supervised learning - simulation programmes: simple perceptrons and steepest-descent learning
- simulation programmes for simple perceptrons
- suppressing noise effets in pattern classifiers and associators - quasilinear perceptrons
- basis-function-expansion layers and nonlinear regression. Part 5 Feedforward multilayer networks and backpropagation: computer simulation of backpropagation networks
- application areas and simulation programmes
- other optimization technqiues. Part 6 Unsupervised competitive learning: unsupervised learning - adaptive vector quantization
- computer simulation of competitive learning
- adaptive resonance emulation
- vector quantization with random input
- measurement of pattern statistics. Part 7 More competitive-learning techniques: supervised competitive learning and counterpropagation networks
- dynamic neuron-layer models for competitive learning and pattern normalization
- competitive learning with multiple winners. Part 8 Associative memories using the equilibrium states of feedback networks - stability theorems: some classical associative memories
- bidirectional associative memories
- network stability. Part 9 Data compression in hidden layers and principal components - vector-shift operations, image processing and miscellaneous topics: data compression layers abstract significant pattern features
- index-shifting neuron layers, vector convolutions and filters
- operations on two-dimensional image patterns
- miscellaneous topics. Part 10 Differential-equation models and time-history patterns - combined simulation of neural networks and other dynamic systems: models with differential equations
- temporal-pattern recognition
- temporal-pattern recognition, time-series prediction and dynamic-model matching. Part 11 Simulation of fuzzy logic, radial-basis-function networks and fuzzy-basis-function networks: fuzzy logic, prediction and control
- fuzzy-logic operations for controller/predictor design
- how to do it - computer simulation of fuzzy-logic control
- radial-basis-function networks
- fuzzy-basis-function networks.
by "Nielsen BookData"
