May, 2022
协方差矩阵适应 MAP 退火
Covariance Matrix Adaptation MAP-Annealing
Matthew C. Fontaine, Stefanos Nikolaidis
TL;DRCMA-MAE solves limitations encountered by QD algorithms by exploring flat objectives and improving performance for low-resolution archives, achieving state-of-the-art performance through its focus on quality diversity and objective exploration.