PREDICTING AYO GAME STRATEGY USING NEURAL NETWORK-BASED REFINEMENT

In this book, the basic neural network technology was discussed as a methodical paradigm in building artificial intelligence. This artificial intelligence is compared with the naturally occurring one. Learning as a process of knowledge acquisition was enunciated and its major types illustrated. Single layer and multilayer model of neural network was outlined. Mention was made of neural network application areas in the past, present and future. As a sample, the eXclusive-OR (XOR) network was modeled and implemented. Ayo otherwise known as Awale or Awari, belongs to the family of board games called Mancala. It is a count-and-capture, two-persons-zero-sum strategic game. The board has twelve pits, six pits on each side and forty-eight seeds in all. As the game commences, the seeds are distributed in fours in each pit. The goal is that one player capture more seeds than the other using better strategies. The Ayo game strategies was modeled and implemented too. All implementations was in C language as a proof of concept. Varying experimental tests and results were discussed.