Quantifying the Impact of Non-Stationarity in Reinforcement Learning-Based Traffic Signal Control
This addresses non-stationarity in traffic optimization, an incremental study focusing on specific effects in a multi-agent RL setting.
The paper tackled the problem of non-stationarity in reinforcement learning for traffic signal control by analyzing how different sources, such as changing traffic patterns and partial observability, impact performance, showing that partial observability causes more drastic performance degradation than changes in traffic patterns.
In reinforcement learning (RL), dealing with non-stationarity is a challenging issue. However, some domains such as traffic optimization are inherently non-stationary. Causes for and effects of this are manifold. In particular, when dealing with traffic signal controls, addressing non-stationarity is key since traffic conditions change over time and as a function of traffic control decisions taken in other parts of a network. In this paper we analyze the effects that different sources of non-stationarity have in a network of traffic signals, in which each signal is modeled as a learning agent. More precisely, we study both the effects of changing the \textit{context} in which an agent learns (e.g., a change in flow rates experienced by it), as well as the effects of reducing agent observability of the true environment state. Partial observability may cause distinct states (in which distinct actions are optimal) to be seen as the same by the traffic signal agents. This, in turn, may lead to sub-optimal performance. We show that the lack of suitable sensors to provide a representative observation of the real state seems to affect the performance more drastically than the changes to the underlying traffic patterns.