Solving equations by hybrid evolutionary computation techniques

Evolutionary computation techniques have mostly been used to solve various optimization and learning problems. This paper describes a novel application of evolutionary computation techniques to equation solving. Several combinations of evolutionary computation techniques and classical numerical methods are proposed to solve linear and partial differential equations. The hybrid algorithms have been compared with the well-known classical numerical methods. The experimental results show that the proposed hybrid algorithms outperform the classical numerical methods significantly in terms of effectiveness and efficiency.