论文标题
人口控制符合DOOB的Martingale定理：无噪声的多模式案例
Population Control meets Doob's Martingale Theorems: the Noise-free Multimodal Case
论文作者
论文摘要
我们研究了一种基于测试的人群大小适应（TBPSA）方法，灵感来自人口控制，在无噪声的多模式情况下。在嘈杂的环境中，TBPSA通常建议在运行结束时，将高斯的中心作为最佳的近似值。我们表明，加上更为幼稚的建议，即推荐到目前为止具有最佳健身价值的访问点，TBPSA在无噪声的多峰环境中也很强大。 我们通过实验证明了这一机制，并从理论上探索了这种机制：我们证明TBPSA能够逃脱高原，尽管它可以收敛到局部最小值。这会导致在多模式设置中有效的算法，而无需从头开始进行随机重新启动。
We study a test-based population size adaptation (TBPSA) method, inspired from population control, in the noise-free multimodal case. In the noisy setting, TBPSA usually recommends, at the end of the run, the center of the Gaussian as an approximation of the optimum. We show that combined with a more naive recommendation, namely recommending the visited point which had the best fitness value so far, TBPSA is also powerful in the noise-free multimodal context. We demonstrate this experimentally and explore this mechanism theoretically: we prove that TBPSA is able to escape plateaus with probability one in spite of the fact that it can converge to local minima. This leads to an algorithm effective in the multimodal setting without resorting to a random restart from scratch.
