We propose a new approach to apply the chaining technique in conjunction with information-theoretic measures to bound the generalization error of machine learning algorithms. Different from the deterministic chaining approach based on hierarchical partitions of a metric space, previously proposed by Asadi et al., we propose a stochastic chaining approach, which replaces the hierarchical partitions with an abstracted Markovian model borrowed from successive refinement source coding. This approach has three benefits over deterministic chaining: 1) the metric space is not necessarily bounded, 2) facilitation of subsequent analysis to yield more explicit bound, and 3) further opportunity to optimize the bound by removing the geometric rigidity of the partitions. The proposed approach includes the traditional chaining as a special case, and can therefore also utilize any deterministic chaining construction. We illustrate these benefits using the problem of estimating Gaussian mean and that of phase retrieval. For the former, we derive a bound that provides an order-wise improvement over previous results, and for the latter we provide a stochastic chain that allows optimization over the chaining parameter.
翻译:我们建议采用一种新的方法,结合信息理论措施应用链锁技术,以约束机器学习算法的普遍错误。不同于Asadi等人先前提出的基于公制空间等级分割的确定性链锁方法,我们建议采用随机链锁方法,用从连续改进源代码中借用的抽象的Markovian模型取代等级分割。这一方法对确定性链有三种好处:(1) 测量空间不一定受约束;(2) 便利随后的分析,以产生更明确的约束;(3) 通过消除隔段的几何僵硬性来优化界限的进一步机会。拟议方法将传统链链条作为特例,因此也可以使用任何确定性链条构造。我们用估算高斯人平均值和阶段检索值的问题来说明这些好处。对于前者,我们得到的界限提供了比以往结果有条理的改进,对于后者,我们提供了一条能够对链链条进行优化的链条。