Title :
A RAM-based neural net with inhibitory weights and its application to recognising handwritten digits
Author :
Jørgensen, Thomas Martini
Author_Institution :
Riso Nat. Lab., Roskilde, Denmark
Abstract :
A method for introducing inhibitory weights into RAM based nets has been developed. The inhibitory weights leads to a more robust net and much lower error rates can be obtained. In the paper we describe how the inhibition factors can be learned with a one shot learning scheme. The main strategy is to choose the inhibitory values so that they minimise the error-rate obtained in a crossvalidating test performed on the training set. The inhibition technique has been tested on the task of recognising handwritten digits. The results obtained match the best error rates reported in the literature
Keywords :
learning (artificial intelligence); neural nets; optical character recognition; random-access storage; RAM-based neural net; crossvalidating test; error rate minimisation; handwritten digit recognition; inhibitory weights; one-shot learning scheme; Error analysis; Handwriting recognition; Laboratories; Neural networks; Performance evaluation; Random access memory; Read-write memory; Robustness; Table lookup; Testing;
Conference_Titel :
Neural Networks for Identification, Control, Robotics, and Signal/Image Processing, 1996. Proceedings., International Workshop on
Conference_Location :
Venice
Print_ISBN :
0-8186-7456-3
DOI :
10.1109/NICRSP.1996.542764
