Abstract
In this paper, a new algorithm of handwritten character recognition based on result-feedback is proposed. It is designed as an effective neural network by adding confidence back-propagation and input modification, thus both pre-processing and recognition operations are closely integrated together. The convergence of the algorithm is proved and many experiments show that the error rate in such a result-feedback neural network (RFNN) can be greatly reduced as well as the robust to environmental noise.
| Original language | English |
|---|---|
| Pages (from-to) | 139-150 |
| Number of pages | 12 |
| Journal | Neural, parallel & scientific computations |
| Volume | 9 |
| Issue number | 2 |
| Publication status | Published - 2001 |
ASJC Scopus subject areas
- Artificial Intelligence
- Computer Networks and Communications
- Software
- Applied Mathematics
- Theoretical Computer Science