Artificial neural network guessing method and game

摘要

A method of operating a neural network to provide a method for guessing an object that the user of an electronic game is thinking of is disclosed. The neural network is structured in a target objects-by-questions matrix and each cell of the matrix defines an input-output connection weight. The network can be operated in two modes, the first mode uses answers to previously asked questions as the input nodes and the target objects as the output nodes, while the second mode has the target objects as the input nodes and the questions as the output nodes. The method involves ranking the target objects by using the network in the first mode, ranking the questions by using the network in the second mode, and providing a guess according to the ranking of the target objects.