Time-Aware Q-Networks: Resolving Temporal Irregularity for Deep Reinforcement Learning

TL;DR

This paper introduces Time-aware Q-Networks (TQN), a deep reinforcement learning framework that explicitly incorporates irregular time intervals between events to improve decision-making in real-world temporal sequences.

Contribution

The work presents a novel TQN framework that considers elapsed time and future action windows, significantly enhancing performance over traditional DQN in irregular time interval scenarios.

Findings

TQN outperforms DQN in various irregular time interval tasks.

Time-aware discounting is crucial in classic RL tasks.

Combining boosting methods with TQN improves learning in real-world applications.

Abstract

Deep Reinforcement Learning (DRL) has shown outstanding performance on inducing effective action policies that maximize expected long-term return on many complex tasks. Much of DRL work has been focused on sequences of events with discrete time steps and ignores the irregular time intervals between consecutive events. Given that in many real-world domains, data often consists of temporal sequences with irregular time intervals, and it is important to consider the time intervals between temporal events to capture latent progressive patterns of states. In this work, we present a general Time-Aware RL framework: Time-aware Q-Networks (TQN), which takes into account physical time intervals within a deep RL framework. TQN deals with time irregularity from two aspects: 1) elapsed time in the past and an expected next observation time for time-aware state approximation, and 2) action time window for the future for time-aware discounting of rewards. Experimental results show that by capturing the underlying structures in the sequences with time irregularities from both aspects, TQNs significantly outperform DQN in four types of contexts with irregular time intervals. More specifically, our results show that in classic RL tasks such as CartPole and MountainCar and Atari benchmark with randomly segmented time intervals, time-aware discounting alone is more important while in the real-world tasks such as nuclear reactor operation and septic patient treatment with intrinsic time intervals, both time-aware state and time-aware discounting are crucial. Moreover, to improve the agent's learning capacity, we explored three boosting methods: Double networks, Dueling networks, and Prioritized Experience Replay, and our results show that for the two real-world tasks, combining all three boosting methods with TQN is especially effective.