Abstract
This paper discusses computer simulations and experimental designs for assessing the binary pattern classification capabilities of basic types of optical neural Al processor. After learning binary input patterns whose binary class is known, these processors classify patterns whose class is unknown. This classification task may be difficult if the relationship between pattern and class has no concise statement or algorithm, as in the case of random problems.1
© 1987 Optical Society of America
PDF ArticleMore Like This
Myung Soo Kim and Clark C. Guest
MJJ3 OSA Annual Meeting (FIO) 1990
Gordon R. Little and Steven C. Gustafson
MD3 OSA Annual Meeting (FIO) 1986
R. D. Griffin, J. N. Lee, T. Maxwell, and F. P. Pursel
FB4 OSA Annual Meeting (FIO) 1987