Solving relative reduction problem using genetic algorithms

Kind of knowledge relative reduction algorithm was proposed. With decision attribute supporting degree applied in knowledge expression system, the support degree of the knowledge, supplied by condition attribute for the whole decision making was described and then relative core was obtained. The relative core was introduced in initial population in GA which can accelerate convergence speed. Penalty function was used in fitness function to assure the reduction has fewer attributes and stronger supporting degree at the same time. The searching effect is very well. The practical simulation results show that the approach was effective in solving knowledge reduction problem.