Cellular Atrophy for Realistic Neural Nets
Computing Sciences and Computer Engineering
The study of memory loss has been going on for a long time. Memories systematically distort and modify with time. Many biological organs go through a continuum of cellular atrophy and regeneration. In this continuum, many of the processing elements will eventually atrophy, reducing the total number of processing elements, thus increasing the system's efficiency. In this paper, we briefly describe memory loss in biological systems and how to present this concept in realistic neural nets, for instance, geometric pattern recognition.
Computers and Industrial Engineering
Ali, D. L.,
(1996). Cellular Atrophy for Realistic Neural Nets. Computers and Industrial Engineering, 31(41276), 435-438.
Available at: https://aquila.usm.edu/fac_pubs/5714