An adaptive evolutionary algorithm based on typical chess problems for tuning a chess evaluation function

This paper presents a method for adjusting weights of the evaluation function of a chess engine. Such an adjustment is carried out through an evolutionary algorithm which adopts a mechanism that selects the virtual players (individuals in the population) that have the highest number of problems properly solved from a database of typical chess problems. This method has the advantage that we only mutate those weights involved in the solution of the current problem. Furthermore, the mutation mechanism is adapted through the number of problems solved by each virtual player. Our results indicate that the material values obtained by our approach are similar to the values known from chess theory. Additionally, we also show that, using the approach proposed here, the strength of our chess engine is increased in 335 points.