Active Learning for Multi-Objective Optimization

Marcela Zuluaga, Guillaume Sergent, Andreas Krause, Markus Püschel
Proceedings of the 30th International Conference on Machine Learning, PMLR 28(1):462-470, 2013.

Abstract

In many fields one encounters the challenge of identifying, out of a pool of possible designs, those that simultaneously optimize multiple objectives. This means that usually there is not one optimal design but an entire set of Pareto-optimal ones with optimal tradeoffs in the objectives. In many applications, evaluating one design is expensive; thus, an exhaustive search for the Pareto-optimal set is unfeasible. To address this challenge, we propose the Pareto Active Learning (PAL) algorithm, which intelligently samples the design space to predict the Pareto-optimal set. Key features of PAL include (1) modeling the objectives as samples 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 sampling cost. We provide theoretical bounds on PAL’s sampling cost required to achieve a desired accuracy. Further, we show an experimental evaluation on three real-world data sets. The results show PAL’s effectiveness; in particular it improves significantly over a state-of-the-art evolutionary algorithm, saving in many cases about 33%.

Cite this Paper


BibTeX
@InProceedings{pmlr-v28-zuluaga13, title = {Active Learning for Multi-Objective Optimization}, author = {Zuluaga, Marcela and Sergent, Guillaume and Krause, Andreas and Püschel, Markus}, booktitle = {Proceedings of the 30th International Conference on Machine Learning}, pages = {462--470}, year = {2013}, editor = {Dasgupta, Sanjoy and McAllester, David}, volume = {28}, number = {1}, series = {Proceedings of Machine Learning Research}, address = {Atlanta, Georgia, USA}, month = {17--19 Jun}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v28/zuluaga13.pdf}, url = {https://proceedings.mlr.press/v28/zuluaga13.html}, abstract = {In many fields one encounters the challenge of identifying, out of a pool of possible designs, those that simultaneously optimize multiple objectives. This means that usually there is not one optimal design but an entire set of Pareto-optimal ones with optimal tradeoffs in the objectives. In many applications, evaluating one design is expensive; thus, an exhaustive search for the Pareto-optimal set is unfeasible. To address this challenge, we propose the Pareto Active Learning (PAL) algorithm, which intelligently samples the design space to predict the Pareto-optimal set. Key features of PAL include (1) modeling the objectives as samples 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 sampling cost. We provide theoretical bounds on PAL’s sampling cost required to achieve a desired accuracy. Further, we show an experimental evaluation on three real-world data sets. The results show PAL’s effectiveness; in particular it improves significantly over a state-of-the-art evolutionary algorithm, saving in many cases about 33%.} }
Endnote
%0 Conference Paper %T Active Learning for Multi-Objective Optimization %A Marcela Zuluaga %A Guillaume Sergent %A Andreas Krause %A Markus Püschel %B Proceedings of the 30th International Conference on Machine Learning %C Proceedings of Machine Learning Research %D 2013 %E Sanjoy Dasgupta %E David McAllester %F pmlr-v28-zuluaga13 %I PMLR %P 462--470 %U https://proceedings.mlr.press/v28/zuluaga13.html %V 28 %N 1 %X In many fields one encounters the challenge of identifying, out of a pool of possible designs, those that simultaneously optimize multiple objectives. This means that usually there is not one optimal design but an entire set of Pareto-optimal ones with optimal tradeoffs in the objectives. In many applications, evaluating one design is expensive; thus, an exhaustive search for the Pareto-optimal set is unfeasible. To address this challenge, we propose the Pareto Active Learning (PAL) algorithm, which intelligently samples the design space to predict the Pareto-optimal set. Key features of PAL include (1) modeling the objectives as samples 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 sampling cost. We provide theoretical bounds on PAL’s sampling cost required to achieve a desired accuracy. Further, we show an experimental evaluation on three real-world data sets. The results show PAL’s effectiveness; in particular it improves significantly over a state-of-the-art evolutionary algorithm, saving in many cases about 33%.
RIS
TY - CPAPER TI - Active Learning for Multi-Objective Optimization AU - Marcela Zuluaga AU - Guillaume Sergent AU - Andreas Krause AU - Markus Püschel BT - Proceedings of the 30th International Conference on Machine Learning DA - 2013/02/13 ED - Sanjoy Dasgupta ED - David McAllester ID - pmlr-v28-zuluaga13 PB - PMLR DP - Proceedings of Machine Learning Research VL - 28 IS - 1 SP - 462 EP - 470 L1 - http://proceedings.mlr.press/v28/zuluaga13.pdf UR - https://proceedings.mlr.press/v28/zuluaga13.html AB - In many fields one encounters the challenge of identifying, out of a pool of possible designs, those that simultaneously optimize multiple objectives. This means that usually there is not one optimal design but an entire set of Pareto-optimal ones with optimal tradeoffs in the objectives. In many applications, evaluating one design is expensive; thus, an exhaustive search for the Pareto-optimal set is unfeasible. To address this challenge, we propose the Pareto Active Learning (PAL) algorithm, which intelligently samples the design space to predict the Pareto-optimal set. Key features of PAL include (1) modeling the objectives as samples 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 sampling cost. We provide theoretical bounds on PAL’s sampling cost required to achieve a desired accuracy. Further, we show an experimental evaluation on three real-world data sets. The results show PAL’s effectiveness; in particular it improves significantly over a state-of-the-art evolutionary algorithm, saving in many cases about 33%. ER -
APA
Zuluaga, M., Sergent, G., Krause, A. & Püschel, M.. (2013). Active Learning for Multi-Objective Optimization. Proceedings of the 30th International Conference on Machine Learning, in Proceedings of Machine Learning Research 28(1):462-470 Available from https://proceedings.mlr.press/v28/zuluaga13.html.

Related Material