Deep Reinforcement Learning for Traffic Signal Control: A Review

Rasheed, Faizan; Yau, Kok‐Lim Alvin; Noor, Rafidah Md; Wu, Celimuge; Low, Yeh Ching

doi:10.1109/access.2020.3034141

Cited by 89 publications

(39 citation statements)

References 69 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Subsequently, it adjusts the traffic phase split based on the traffic condition [6]. The fully-dynamic traffic light controllers are more realistic in monitoring traffic movements because it uses longer-term traffic conditions, and the approach has shown to alleviate traffic congestion more effectively compared to deterministic and semi-dynamic traffic light controllers [7][8][9].…”

Section: Types Of Traffic Light Controllersmentioning

confidence: 99%

Deep Reinforcement Learning for Addressing Disruptions in Traffic Light Control

Rasheed¹,

Yau²,

Noor³

et al. 2022

Computers, Materials &Amp; Continua

Self Cite

View full text Add to dashboard Cite

This paper investigates the use of multi-agent deep Q-network (MADQN) to address the curse of dimensionality issue occurred in the traditional multi-agent reinforcement learning (MARL) approach. The proposed MADQN is applied to traffic light controllers at multiple intersections with busy traffic and traffic disruptions, particularly rainfall. MADQN is based on deep Q-network (DQN), which is an integration of the traditional reinforcement learning (RL) and the newly emerging deep learning (DL) approaches. MADQN enables traffic light controllers to learn, exchange knowledge with neighboring agents, and select optimal joint actions in a collaborative manner. A case study based on a real traffic network is conducted as part of a sustainable urban city project in the Sunway City of Kuala Lumpur in Malaysia. Investigation is also performed using a grid traffic network (GTN) to understand that the proposed scheme is effective in a traditional traffic network. Our proposed scheme is evaluated using two simulation tools, namely Matlab and Simulation of Urban Mobility (SUMO). Our proposed scheme has shown that the cumulative delay of vehicles can be reduced by up to 30% in the simulations.

show abstract

Section: Types Of Traffic Light Controllersmentioning

confidence: 99%

Deep Reinforcement Learning for Addressing Disruptions in Traffic Light Control

Rasheed¹,

Yau²,

Noor³

et al. 2022

Computers, Materials &Amp; Continua

Self Cite

View full text Add to dashboard Cite

show abstract

“…Meanwhile, deep learning is a relative new learning paradigm that integrates the multilayer perceptron approach, which consists of a large number of layers of neurons, into supervised and reinforcement learning approaches. Such integration has shown to address the shortcomings of the original learning paradigms [40]. Overall, further investigation can be pursued to explore and exploit the use of the reinforcement learning and deep learning approaches since the need for human effort to categorize data using labels has become a mammoth task with big data.…”

Section: Learning Paradigmsmentioning

confidence: 99%

Artificial Intelligence Marketing (AIM) for Enhancing Customer Relationships

2021

Self Cite

View full text Add to dashboard Cite

Based on the literature, we present an artificial intelligence marketing (AIM) framework that enables autonomous machines to receive big data and information, use artificial intelligence (AI) to create knowledge, and then disseminate and apply the knowledge to enhance customer relationships in a knowledge-based environment. To develop the AIM framework, we bring together and curate a wide range of relevant literatures including real-life examples and cases, and then understand how these literatures contribute to the framework in this research topic. We explain the AIM framework from the interdisciplinary perspective, which is an important role of both the artificial intelligence and marketing academia. The AIM framework includes three main components, including the pre-processor, the main processor, and the memory storage. The main processor, which is the key component, uses AI to process structured data processed by pre-processor in order to make real-time decisions and reasonings. The AI approach is characterized by its hypothetical abilities, learning paradigms, and operation modes with human. The strategic use of the developed AIM framework based on the literature to enhance customer relationships, including customer trust, satisfaction, commitment, engagement, and loyalty, is presented. Finally, future potential investigations are presented to drive forward this interdisciplinary research topic.

show abstract

“…Multiple agents exchange knowledge among themselves in order to achieve global optimization in a collaborative manner, and so it addresses the moving target problem. Knowledge exchange enables agents to consider their own and neighboring agents' performances [32].…”

Section: Multi-agent Deep Q-networkmentioning

confidence: 99%

Applications of Multi-Agent Deep Reinforcement Learning: Models and Algorithms

et al. 2021

Self Cite

View full text Add to dashboard Cite

Recent advancements in deep reinforcement learning (DRL) have led to its application in multi-agent scenarios to solve complex real-world problems, such as network resource allocation and sharing, network routing, and traffic signal controls. Multi-agent DRL (MADRL) enables multiple agents to interact with each other and with their operating environment, and learn without the need for external critics (or teachers), thereby solving complex problems. Significant performance enhancements brought about by the use of MADRL have been reported in multi-agent domains; for instance, it has been shown to provide higher quality of service (QoS) in network resource allocation and sharing. This paper presents a survey of MADRL models that have been proposed for various kinds of multi-agent domains, in a taxonomic approach that highlights various aspects of MADRL models and applications, including objectives, characteristics, challenges, applications, and performance measures. Furthermore, we present open issues and future directions of MADRL.

show abstract

Deep Reinforcement Learning for Traffic Signal Control: A Review

Cited by 89 publications

References 69 publications

Deep Reinforcement Learning for Addressing Disruptions in Traffic Light Control

Deep Reinforcement Learning for Addressing Disruptions in Traffic Light Control

Artificial Intelligence Marketing (AIM) for Enhancing Customer Relationships

Applications of Multi-Agent Deep Reinforcement Learning: Models and Algorithms

Contact Info

Product

Resources

About