Reward-Free Model-Based Reinforcement Learning with Linear Function Approximation

by   Weitong Zhang, et al.

We study the model-based reward-free reinforcement learning with linear function approximation for episodic Markov decision processes (MDPs). In this setting, the agent works in two phases. In the exploration phase, the agent interacts with the environment and collects samples without the reward. In the planning phase, the agent is given a specific reward function and uses samples collected from the exploration phase to learn a good policy. We propose a new provably efficient algorithm, called UCRL-RFE under the Linear Mixture MDP assumption, where the transition probability kernel of the MDP can be parameterized by a linear function over certain feature mappings defined on the triplet of state, action, and next state. We show that to obtain an ϵ-optimal policy for arbitrary reward function, UCRL-RFE needs to sample at most Õ(H^5d^2ϵ^-2) episodes during the exploration phase. Here, H is the length of the episode, d is the dimension of the feature mapping. We also propose a variant of UCRL-RFE using Bernstein-type bonus and show that it needs to sample at most Õ(H^4d(H + d)ϵ^-2) to achieve an ϵ-optimal policy. By constructing a special class of linear Mixture MDPs, we also prove that for any reward-free algorithm, it needs to sample at least Ω̃(H^2dϵ^-2) episodes to obtain an ϵ-optimal policy. Our upper bound matches the lower bound in terms of the dependence on ϵ and the dependence on d if H ≥ d.


page 1

page 2

page 3

page 4


Optimal Horizon-Free Reward-Free Exploration for Linear Mixture MDPs

We study reward-free reinforcement learning (RL) with linear function ap...

Reward-Free RL is No Harder Than Reward-Aware RL in Linear Markov Decision Processes

Reward-free reinforcement learning (RL) considers the setting where the ...

Adaptive Reward-Free Exploration

Reward-free exploration is a reinforcement learning setting recently stu...

Self-Supervised Online Reward Shaping in Sparse-Reward Environments

We propose a novel reinforcement learning framework that performs self-s...

Near-Optimal Reward-Free Exploration for Linear Mixture MDPs with Plug-in Solver

Although model-based reinforcement learning (RL) approaches are consider...

Provably Efficient Reward-Agnostic Navigation with Linear Value Iteration

There has been growing progress on theoretical analyses for provably eff...

Learning One Representation to Optimize All Rewards

We introduce the forward-backward (FB) representation of the dynamics of...

Please sign up or login with your details

Forgot password? Click here to reset