Modified Neural Net for the Boolean Satisfiability Problem
Manuel Mejía-Lavalle, A. Jose Ruiz, O. Joaquín Pérez, Sánchez Cervantes · 2015
A modified Hopfield Artificial Neural Network is proposed to solve effectively and efficiently Boolean Satisfiability (SAT) NP-hard problems. The proposed Neural Network is compared against other traditional methods employed in this field, such as Greedy SAT and Genetic Algorithms for SAT. The results show that the proposed network represents a good alternative given their output quality and response time speed.