Memory system for use of modeling psychological functions of the brain

A method of forming an artificial neural network structure modeling psychological functions of a brain, said method comprising the steps of:

• forming a short-term memory neural network on the basis of input pattern vectors;
• forming a long-term memory neural network for ordinary outputs on the basis of output vectors provided by the short-term memory neural network;
• and
• successively forming hierarchical long-term memory neural networks for ordinary outputs by using parameters of neurons in the formed long-term memory neural network for ordinary outputs.

A memory-chaining system having an artificial neural network structure modeling memory-chaining functions of a brain, comprising:

• a long-term memory network unit using an artificial neural network structure including a plurality of neurons;
• and
• a network control unit that connects the neurons of the long-term memory network unit with each other to realize a memory-chaining function;
• wherein each of the neurons of the long-term memory network unit has a radial-basis function main part that provides an activation intensity corresponding to similarity between an input vector and a centroid vector according to a radial-basis function;
• and a pointer part that holds pointer information about a mutually related other neuron.