论文标题
筛选规则及其在主动设置标识的复杂性
Screening Rules and its Complexity for Active Set Identification
论文作者
论文摘要
最近引入了筛选规则，作为一种在机器学习中出现的优化问题中明确识别主动结构（例如稀疏性）的技术。这导致了基于大幅度降低的新加速方法。我们表明，筛查规则源于细分集和最佳条件的自然特性的结合，因此可以以统一的方式理解。在温和的假设下，我们分析了确定任何收敛算法的最佳活动集所需的迭代次数。我们表明，这仅取决于其收敛率。
Screening rules were recently introduced as a technique for explicitly identifying active structures such as sparsity, in optimization problem arising in machine learning. This has led to new methods of acceleration based on a substantial dimension reduction. We show that screening rules stem from a combination of natural properties of subdifferential sets and optimality conditions, and can hence be understood in a unified way. Under mild assumptions, we analyze the number of iterations needed to identify the optimal active set for any converging algorithm. We show that it only depends on its convergence rate.
