论文标题
关于在进化计算中平均最佳样本
On averaging the best samples in evolutionary computation
论文作者
论文摘要
在进化计算中，选择正确的选择率是一个长期存在的问题。在连续不受约束的情况下，我们从数学上证明了单亲$μ= 1 $会导致球函数的情况下的简单遗憾。我们提供理论上的选择率$μ/λ$，从而提高进度率。有了我们选择的选择率，我们将获得订单$ O（λ^{ - 1}）$的可证明后悔，该$必须与$ o（λ^{ - 2/d}）$相比，在$μ= 1 $的情况下。我们通过实验完成研究以确认我们的理论主张。
Choosing the right selection rate is a long standing issue in evolutionary computation. In the continuous unconstrained case, we prove mathematically that a single parent $μ=1$ leads to a sub-optimal simple regret in the case of the sphere function. We provide a theoretically-based selection rate $μ/λ$ that leads to better progress rates. With our choice of selection rate, we get a provable regret of order $O(λ^{-1})$ which has to be compared with $O(λ^{-2/d})$ in the case where $μ=1$. We complete our study with experiments to confirm our theoretical claims.
