In many fields one encounters the challenge of identifying out of a pool of possible designs those that simultaneously optimize multiple objectives. In many applications an exhaustive search for the Pareto-optimal set is infeasible. To address this challenge, we propose the ε-Pareto Active Learning (ε-PAL) algorithm which adaptively samples the design space to predict a set of Pareto-optimal solutions that cover the true Pareto front of the design space with some granularity regulated by a parameter ε. Key features of ε-PAL include (1) modeling the objectives as draws from a Gaussian process distribution to capture structure and accommodate noisy evaluation; (2) a method to carefully choose the next design to evaluate to maximize progress; and (3) the ability to control prediction accuracy and sam- pling cost. We provide theoretical bounds on ε-PAL's sampling cost required to achieve a desired accuracy. Further, we perform an experimental evaluation on three real-world data sets that demonstrate ε-PAL's effectiveness; in comparison to the state-of-the-art active learning algorithm PAL, ε-PAL reduces the amount of computations and the number of samples from the design space required to meet the user's desired level of accuracy. In addition, we show that ε-PAL improves significantly over a state-of-the-art multi-objective optimization method, saving in most cases 30% to 70% evaluations to achieve the same accuracy.

@article{zuluaga16active,
	author = {Marcela Zuluaga and Andreas Krause and Markus P{\"u}schel},
	journal = {Journal of Machine Learning Research (JMLR; accepted for publication)},
	month = {August},
	number = {104},
	pages = {1−32},
	title = {ε-PAL: An Active Learning Approach to the Multi-Objective Optimization Problem},
	volume = {17},
	year = {2016}}