In many real world problems, we want to infer some property of an expensive black-box function f, given a budget of T function evaluations. One example is budget constrained global optimization of f, for which Bayesian optimization is a popular method. Other properties of interest include local optima, level sets, integrals, or graph-structured information induced by f. Often, we can find an algorithm A to compute the desired property, but it may require far more than T queries to execute. Given such an A, and a prior distribution over f, we refer to the problem of inferring the output of A using T evaluations as Bayesian Algorithm Execution (BAX). To tackle this problem, we present a procedure, InfoBAX, that sequentially chooses queries that maximize mutual information with respect to the algorithm's output. Applying this to Dijkstra's algorithm, for instance, we infer shortest paths in synthetic and real-world graphs with black-box edge costs. Using evolution strategies, we yield variants of Bayesian optimization that target local, rather than global, optima. On these problems, InfoBAX uses up to 500 times fewer queries to f than required by the original algorithm. Our method is closely connected to other Bayesian optimal experimental design procedures such as entropy search methods and optimal sensor placement using Gaussian processes.
翻译:在许多真实的世界问题中, 我们想要根据 T 函数评价的预算来推断昂贵的黑盒函数 f 的某些属性。 一个例子是预算限制的f 全球优化, 巴伊西亚优化是一种受欢迎的方法。 其他感兴趣的属性包括本地的optima、 级别设置、 集成集、 或图形结构信息 。 通常, 我们可以找到一个算法 A 来计算想要的属性, 但是它可能比要执行的查询要多得多 。 鉴于此 A 和 之前的分布, 我们指的是使用 Bayesian Algorithm 执行( BAX ) 来推断 A 的输出。 为了解决这个问题, 我们提出了一个程序, 即 InfoBAX 。 该程序依次选择查询, 使对算法输出的相互信息最大化 。 将此应用到 Dijkstra 的算法, 但它可能比 需要执行的问法要短得多得多 。 使用进化战略, 我们给出了 Bayes 优化的方法是本地的, 而不是全球的, 选择 。 在这些问题上, InBOX 使用比原始的原始的搜索方法要少一些。