Recomposing the Reinforcement Learning Building Blocks with Hypernetworks

Elad Sarafian, Shai Keynan, Sarit Kraus

نتاج البحث: فصل من :كتاب / تقرير / مؤتمرمنشور من مؤتمرمراجعة النظراء

ملخص

The Reinforcement Learning (RL) building blocks, i.e. Q-functions and policy networks, usually take elements from the cartesian product of two domains as input. In particular, the input of the Q-function is both the state and the action, and in multi-task problems (Meta-RL) the policy can take a state and a context. Standard architectures tend to ignore these variables' underlying interpretations and simply concatenate their features into a single vector. In this work, we argue that this choice may lead to poor gradient estimation in actor-critic algorithms and high variance learning steps in Meta-RL algorithms. To consider the interaction between the input variables, we suggest using a Hypernetwork architecture where a primary network determines the weights of a conditional dynamic network. We show that this approach improves the gradient approximation and reduces the learning step variance, which both accelerates learning and improves the final performance. We demonstrate a consistent improvement across different locomotion tasks and different algorithms both in RL (TD3 and SAC) and in Meta-RL (MAML and PEARL).

اللغة الأصليةالإنجليزيّة
عنوان منشور المضيفProceedings of the 38th International Conference on Machine Learning, ICML 2021
الصفحات9301-9312
عدد الصفحات12
رقم المعيار الدولي للكتب (الإلكتروني)9781713845065
حالة النشرنُشِر - 2021
الحدث38th International Conference on Machine Learning, ICML 2021 - Virtual, Online
المدة: ١٨ يوليو ٢٠٢١٢٤ يوليو ٢٠٢١

سلسلة المنشورات

الاسمProceedings of Machine Learning Research
مستوى الصوت139

!!Conference

!!Conference38th International Conference on Machine Learning, ICML 2021
المدينةVirtual, Online
المدة١٨/٠٧/٢١٢٤/٠٧/٢١

All Science Journal Classification (ASJC) codes

  • !!Artificial Intelligence
  • !!Software
  • !!Control and Systems Engineering
  • !!Statistics and Probability

بصمة

أدرس بدقة موضوعات البحث “Recomposing the Reinforcement Learning Building Blocks with Hypernetworks'. فهما يشكلان معًا بصمة فريدة.

قم بذكر هذا