Symbolic logic inference system based on recurrent multilayered perceptron neural networks

A method of implementing symbolic logic inference system using a recurrent multilayered perceptron neural network is presented in this paper. Domain rule knowledge can be either acquired through learning domain sample set by a neural network or encoded into a neural network directly. Once the domain rule knowledge has been stored in a neural network, the neural network can be used to implement any symbolic logic inference of that domain. It is a theoretical base for studying relations between the abstract thought of human (symbolic logic inference) and thinking in images of a neural network (linked numeric calculation)