概要:

Neural network is one of the artificial intelligence devised based on brain of human. And, the chaotic neural network is one of the neural network introduced chaos theory.

The goal of this study is to reproduces the phenomenon seen in clinical case of H. M. and the medium term memory defined based on clinical case of H. M. by connecting two chaotic neural networks. One network is the hippocampus module based on the hippocampus and the other network is the cortex module based on the cortex. Hippocampus module has fast learning speed and can memorize medium term memory that is non persistent. Cortex module has slow learning speed and can memorize long term memory that is persistent. By connecting these chaotic neural networks and constructing a network, it would appear that the neural network becomes more close to the human brain to memorize classifying medium term memory and long term memory and to reproduce anterograde amnesia and partially retrograde amnesia seen in clinical case of H. M.