| 摘要 |
PURPOSE: A neuron circuit network for a common knowledge inference is provided for implementing various flexible probability input and outputs by newly adapting a time-based probability. CONSTITUTION: A plurality of cells and connection strengths are provided. A neuron network in which a probability computation is possible using a method for computing an output of an input data and a method in which an interface is obtained between the cells is constituted. The present invention introduces a common knowledge inference based on a neuron network using a probability. Therefore, it is possible to implement various probability input and outputs by newly adapting a time-based probability. The cells reacts based on the probability. It is possible to obtain a quick search because the neuron network operates in parallel.
|
