Abstract
Various lateral inhibition neural networks are studied by means of computer simulation regarding their ability to sharpen the input excitation curves. To quantify this ability we introduced two new entropy-like quantities (called the iteration entropy and the rate of convergence) which represent suitable measures for the sharpening of the input excitation curves in a certain lateral inhibition neural network. Using these quantities we quantitatively described the sharpening ability of different lateral inhibition networks.
Original language | English |
---|---|
Pages (from-to) | 65-80 |
Number of pages | 16 |
Journal | International Journal of Neural Systems |
Volume | 4 |
Issue number | 1 |
DOIs | |
Publication status | Published - 1 Jan 1993 |
Externally published | Yes |