conftrace_

2018 IJCAI IJCAI 2018

Exploration by Distributional Reinforcement Learning

Abstract

We propose a framework based on distributional reinforcement learning and recent attempts to combine Bayesian parameter updates with deep reinforcement learning. We show that our proposed framework conceptually unifies multiple previous methods in exploration. We also derive a practical algorithm that achieves efficient exploration on challenging control tasks.

🌉 Interdisciplinary Bridge - Artificial Intelligence and Reinforcement Learning

🐣 Hot Topic Early Bird - policy evaluation

Authors

Yunhao Tang, Shipra Agrawal

Topics

Artificial Intelligence > Core AI > Agent Systems Machine Learning > Optimization & Theory > Bayesian Inference Reinforcement Learning > Methods > Deep RL Reinforcement Learning > Applications > Value Iteration Machine Learning > Learning Types > Reinforcement Learning

Keywords

deep reinforcement learning policy evaluation bayesian inference distributional reinforcement learning exploration strategy reward distribution bayesian parameter update

More from IJCAI 2018

All IJCAI 2018 papers →

Positive and Unlabeled Learning via Loss Decomposition and Centroid Estimation

Student-t Variational Autoencoder for Robust Density Estimation

Incomplete Multi-View Weak-Label Learning

Algorithms or Actions? A Study in Large-Scale Reinforcement Learning

Deep into Hypersphere: Robust and Unsupervised Anomaly Discovery in Dynamic Networks

Differentiable Submodular Maximization