## Implementing neural networksUniversity of California, San Diego, Department of Electrical Engineering/Computer Science, 1988 - Computers - 292 pages |

### What people are saying - Write a review

We haven't found any reviews in the usual places.

### Contents

Theoretical Aspects of Back Propagation | 14 |

The Two Wire Analog Back Propagation System | 43 |

Typical Squash Function and Derivative | 59 |

Copyright | |

9 other sections not shown

### Common terms and phrases

activation function approximately assumed assumption Back Propagation algorithm backwards flow boolean functions capacitor chain rule Chapter connected convergence data spikes Discrete Weight algorithm Discrete Weight neural dopj equation error information Error output error signal error value Figures 4.2A Finite Step floating point flow of error forward flow gradient descent Hence hidden layer hidden processing elements inhibitory spike input layer input lines input processing elements learn rate line 82 Linear Threshold Units minima monotonically decreasing multiplication neuron number of bits number of spikes output layer output processing elements output signal output units partial derivative pattern presentation Perceptron performed Poisson distribution presentation of pattern probability problem processing element circuits Pulse Width Modulator Pulsed Error line signal trace single bit weights single weight Spike Model Spike Trains squash function summation synaptic elements systolic array total error traditional Back Propagation utilizing VLSI weight change Weight neural network weight value zero