User´s attention knowledge learning in interactive evolutionary computation

A user´s attention in interactive evolutionary computation (IEC) is an important issue. The methods to learn the user´s attention knowledge in IEC are studied in this paper. Firstly, the definition of the user´s attention is given. Secondly, the user´s attention on gene sense unit and some related theorems are given. Based on these theorems, the methods to learn the user´s attention knowledge are presented. Thirdly, a new method to improve the performance of IEC based on the user´s attention knowledge is given. The experiments validate the efficiency of the methods. The study on the user´s attention in IEC establishes a necessary foundation for reducing users´ fatigue in IEC.