Improving the performance of predicting users' subjective evaluation characteristics to reduce their fatigue in IEC

Abstract

Users' fatigue is the biggest technological hurdle facing Interactive Evolutionary Computation (IEC). This paper introduces the idea of "absolute scale" and "neighbour scale" to improve the performance of predicting users' subjective evaluation characteristics in IEC, and thus it will accelerate EC convergence and reduce users' fatigue. We experimentally evaluate the effect of the proposed method using two benchmark functions. The experimental results show that the convergence speed of IEC using the proposed predictor, which learns from absolute evaluation data, is much faster than the conventional one, which learns from relative data, especially in early generations. Also, IEC with predictors that use recent data are more effective than those which use all past data.