@inproceedings{2023_3C_MetaLMRS,
  title={Learning-to-Learn to Guide Random Search: Derivative-Free Meta Blackbox Optimization on Manifold},
  author={Bilgehan Sel and Ahmad Al-Tawaha and Yuhao Ding and Ruoxi Jia and Bo Ji and Javad Lavaei and  Ming Jin},
  booktitle={Learning for Dynamics & Control Conference (L4DC)},
  note = {<font style="color:#FF0000">(oral presentation)</font>},
  pages={},
  year={2023},
  url_pdf={Meta_LMRS.pdf},
  keywords = {Optimization, Reinforcement learning, Machine Learning},
  abstract={Solving a sequence of high-dimensional, nonconvex, but potentially similar optimization problems poses a computational challenge in engineering applications. We propose the \emph{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.  },
}

{"_id":"98EjidjaDhb7ahbpb","bibbaseid":"sel-altawaha-ding-jia-ji-lavaei-jin-learningtolearntoguiderandomsearchderivativefreemetablackboxoptimizationonmanifold-2023","author_short":["Sel, B.","Al-Tawaha, A.","Ding, Y.","Jia, R.","Ji, B.","Lavaei, J.","Jin, M."],"bibdata":{"bibtype":"inproceedings","type":"inproceedings","title":"Learning-to-Learn to Guide Random Search: Derivative-Free Meta Blackbox Optimization on Manifold","author":[{"firstnames":["Bilgehan"],"propositions":[],"lastnames":["Sel"],"suffixes":[]},{"firstnames":["Ahmad"],"propositions":[],"lastnames":["Al-Tawaha"],"suffixes":[]},{"firstnames":["Yuhao"],"propositions":[],"lastnames":["Ding"],"suffixes":[]},{"firstnames":["Ruoxi"],"propositions":[],"lastnames":["Jia"],"suffixes":[]},{"firstnames":["Bo"],"propositions":[],"lastnames":["Ji"],"suffixes":[]},{"firstnames":["Javad"],"propositions":[],"lastnames":["Lavaei"],"suffixes":[]},{"firstnames":["Ming"],"propositions":[],"lastnames":["Jin"],"suffixes":[]}],"booktitle":"Learning for Dynamics & Control Conference (L4DC)","note":"<font style=\"color:#FF0000\">(oral presentation)</font>","pages":"","year":"2023","url_pdf":"Meta_LMRS.pdf","keywords":"Optimization, Reinforcement learning, Machine Learning","abstract":"Solving a sequence of high-dimensional, nonconvex, but potentially similar optimization problems poses a computational challenge in engineering applications. We propose the \\emphfirst 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. ","bibtex":"@inproceedings{2023_3C_MetaLMRS,\n title={Learning-to-Learn to Guide Random Search: Derivative-Free Meta Blackbox Optimization on Manifold},\n author={Bilgehan Sel and Ahmad Al-Tawaha and Yuhao Ding and Ruoxi Jia and Bo Ji and Javad Lavaei and Ming Jin},\n booktitle={Learning for Dynamics & Control Conference (L4DC)},\n note = {<font style=\"color:#FF0000\">(oral presentation)</font>},\n pages={},\n year={2023},\n url_pdf={Meta_LMRS.pdf},\n keywords = {Optimization, Reinforcement learning, Machine Learning},\n abstract={Solving a sequence of high-dimensional, nonconvex, but potentially similar optimization problems poses a computational challenge in engineering applications. We propose the \\emph{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. },\n}\n\n","author_short":["Sel, B.","Al-Tawaha, A.","Ding, Y.","Jia, R.","Ji, B.","Lavaei, J.","Jin, M."],"key":"2023_3C_MetaLMRS","id":"2023_3C_MetaLMRS","bibbaseid":"sel-altawaha-ding-jia-ji-lavaei-jin-learningtolearntoguiderandomsearchderivativefreemetablackboxoptimizationonmanifold-2023","role":"author","urls":{" pdf":"http://www.jinming.tech/papers/Meta_LMRS.pdf"},"keyword":["Optimization","Reinforcement learning","Machine Learning"],"metadata":{"authorlinks":{}},"downloads":57},"bibtype":"inproceedings","biburl":"http://www.jinming.tech/papers/myref.bib","dataSources":["sTzDHHaipTZWjp8oe","Y64tp2HnDCfXgLdc5"],"keywords":["optimization","reinforcement learning","machine learning"],"search_terms":["learning","learn","guide","random","search","derivative","free","meta","blackbox","optimization","manifold","sel","al-tawaha","ding","jia","ji","lavaei","jin"],"title":"Learning-to-Learn to Guide Random Search: Derivative-Free Meta Blackbox Optimization on Manifold","year":2023,"downloads":57}