✦ LIBER ✦
5517667 Neural network that does not require repetitive training
✍ Scribed by Wang Shay-Ping
- Publisher
- Elsevier Science
- Year
- 1996
- Tongue
- English
- Weight
- 84 KB
- Volume
- 11
- Category
- Article
- ISSN
- 0957-4174
No coin nor oath required. For personal study only.
✦ Synopsis
Method and apparatus of error back-propagation for use in a neural network system. A first group of processing devices performs the resolving steps and a second group of analogous processing devices performs the training steps while backpropagating errors calculated in a central processing device. The synaptic coefficient matrix Cij of the first group and the transposed matrix Tji of the second group are simultaneously updated. This updating of the synaptic coefficients can be performed by means of multipliers and adders.