Deep Q-Network with Predictive State Models in Partially Observable Domains

Abstract

While deep reinforcement learning (DRL) has achieved great success in some large domains, most of the related algorithms assume that the state of the underlying system is fully observable. However, many real-world problems are actually partially observable. For systems with continuous observation, most of the related algorithms, e.g., the deep Q-network (DQN) and deep recurrent Q-network (DRQN), use history observations to represent states; however, they often make computation-expensive and ignore the information of actions. Predictive state representations (PSRs) can offer a powerful framework for modelling partially observable dynamical systems with discrete or continuous state space, which represents the latent state using completely observable actions and observations. In this paper, we present a PSR model-based DQN approach which combines the strengths of the PSR model and DQN planning. We use a recurrent network to establish the recurrent PSR model, which can fully learn dynamics of the partially continuous observable environment. Then, the model is used for the state representation and update of DQN, which makes DQN no longer rely on a fixed number of history observations or recurrent neural network (RNN) to represent states in the case of partially observable environments. The strong performance of the proposed approach is demonstrated on a set of robotic control tasks from OpenAI Gym by comparing with the technique with the memory-based DRQN and the state-of-the-art recurrent predictive state policy (RPSP) networks. Source code is available at https://github.com/RPSR-DQN/paper-code.git.

Autonomous Taxi Driving Environment Using Reinforcement Learning Algorithms

“…To put it another way, To estimate the action value, DQN uses a neural network. The most recent four frames of data are fed to the first layer of DQNs for determining current states which are transformed into vectors of action values in connected layers [39]. Rewards of DQNs are shown in Figure 3.…”

“…The average reward achieved by DQN is 13.0. DQN minimizes a differentiable loss function L(θ) [39] by adjusting the network weights θ to improve the action value function.…”