Learning-to-Learn to Guide Random Search: Derivative-Free Meta Blackbox Optimization on Manifold

Bilgehan Sel, Ahmad Tawaha, Yuhao Ding, Ruoxi Jia, Bo Ji, Javad Lavaei, Ming Jin
Proceedings of The 5th Annual Learning for Dynamics and Control Conference, PMLR 211:38-50, 2023.

Abstract

Solving a sequence of high-dimensional, nonconvex, but potentially similar optimization problems poses a computational challenge in engineering applications. We propose the first meta-learning framework that leverages the shared structure among sequential tasks to improve the computational efficiency and sample complexity of derivative-free optimization. Based on the observation that most practical high-dimensional functions lie on a latent low-dimensional manifold, which can be further shared among instances, our method jointly learns the meta-initialization of a search point and a meta-manifold. Theoretically, we establish the benefit of meta-learning in this challenging setting. Empirically, we demonstrate the effectiveness of the proposed algorithm in two high-dimensional reinforcement learning tasks.

Cite this Paper

BibTeX
@InProceedings{pmlr-v211-sel23a, title = {Learning-to-Learn to Guide Random Search: Derivative-Free Meta Blackbox Optimization on Manifold}, author = {Sel, Bilgehan and Tawaha, Ahmad and Ding, Yuhao and Jia, Ruoxi and Ji, Bo and Lavaei, Javad and Jin, Ming}, booktitle = {Proceedings of The 5th Annual Learning for Dynamics and Control Conference}, pages = {38--50}, year = {2023}, editor = {Matni, Nikolai and Morari, Manfred and Pappas, George J.}, volume = {211}, series = {Proceedings of Machine Learning Research}, month = {15--16 Jun}, publisher = {PMLR}, pdf = {https://proceedings.mlr.press/v211/sel23a/sel23a.pdf}, url = {https://proceedings.mlr.press/v211/sel23a.html}, abstract = {Solving a sequence of high-dimensional, nonconvex, but potentially similar optimization problems poses a computational challenge in engineering applications. We propose the first meta-learning framework that leverages the shared structure among sequential tasks to improve the computational efficiency and sample complexity of derivative-free optimization. Based on the observation that most practical high-dimensional functions lie on a latent low-dimensional manifold, which can be further shared among instances, our method jointly learns the meta-initialization of a search point and a meta-manifold. Theoretically, we establish the benefit of meta-learning in this challenging setting. Empirically, we demonstrate the effectiveness of the proposed algorithm in two high-dimensional reinforcement learning tasks.} }
Endnote
%0 Conference Paper %T Learning-to-Learn to Guide Random Search: Derivative-Free Meta Blackbox Optimization on Manifold %A Bilgehan Sel %A Ahmad Tawaha %A Yuhao Ding %A Ruoxi Jia %A Bo Ji %A Javad Lavaei %A Ming Jin %B Proceedings of The 5th Annual Learning for Dynamics and Control Conference %C Proceedings of Machine Learning Research %D 2023 %E Nikolai Matni %E Manfred Morari %E George J. Pappas %F pmlr-v211-sel23a %I PMLR %P 38--50 %U https://proceedings.mlr.press/v211/sel23a.html %V 211 %X Solving a sequence of high-dimensional, nonconvex, but potentially similar optimization problems poses a computational challenge in engineering applications. We propose the first meta-learning framework that leverages the shared structure among sequential tasks to improve the computational efficiency and sample complexity of derivative-free optimization. Based on the observation that most practical high-dimensional functions lie on a latent low-dimensional manifold, which can be further shared among instances, our method jointly learns the meta-initialization of a search point and a meta-manifold. Theoretically, we establish the benefit of meta-learning in this challenging setting. Empirically, we demonstrate the effectiveness of the proposed algorithm in two high-dimensional reinforcement learning tasks.
APA
Sel, B., Tawaha, A., Ding, Y., Jia, R., Ji, B., Lavaei, J. & Jin, M.. (2023). Learning-to-Learn to Guide Random Search: Derivative-Free Meta Blackbox Optimization on Manifold. Proceedings of The 5th Annual Learning for Dynamics and Control Conference, in Proceedings of Machine Learning Research 211:38-50 Available from https://proceedings.mlr.press/v211/sel23a.html.

Related Material