• Login
    ARAN - Access to Research at NUI Galway
    View Item 
    •   ARAN Home
    • Support Services
    • Externally hosted open access publications with NUI Galway authors (2)
    • View Item
    •   ARAN Home
    • Support Services
    • Externally hosted open access publications with NUI Galway authors (2)
    • View Item
    JavaScript is disabled for your browser. Some features of this site may not work without it.

    Browse

    All of ARANCommunities & CollectionsBy Issue DateAuthorsTitlesSubjectsTypesThis CollectionBy Issue DateAuthorsTitlesSubjectsTypes

    My Account

    LoginRegister

    Statistics

    Most Popular ItemsStatistics by CountryMost Popular Authors

    Help

    How to submit and FAQs

    traffic light control using deep policy-gradient and value-function based reinforcement learning

    Thumbnail
    View/Open
    Full Text
    Date
    2017-08-11
    Author
    Mousavi, Seyed Sajad
    schukat, Michael
    Howley, Enda
    Metadata
    Show full item record
    Usage
    This item's downloads: 0 (view details)
    Cited 69 times in Scopus (view citations)
    
    Recommended Citation
    Mousavi, Seyed Sajad; schukat, Michael; Howley, Enda (2017). traffic light control using deep policy-gradient and value-function based reinforcement learning . IET Intelligent Transport Systems 11 (7), 417-423
    Published Version
    http://arxiv.org/pdf/1704.08883
    Abstract
    Recent advances in combining deep neural network architectures with reinforcement learning (RL) techniques have shown promising potential results in solving complex control problems with high-dimensional state and action spaces. Inspired by these successes, in this study, the authors built two kinds of RL algorithms: deep policy-gradient (PG) and value-function-based agents which can predict the best possible traffic signal for a traffic intersection. At each time step, these adaptive traffic light control agents receive a snapshot of the current state of a graphical traffic simulator and produce control signals. The PG-based agent maps its observation directly to the control signal; however, the value-function-based agent first estimates values for all legal control signals. The agent then selects the optimal control action with the highest value. Their methods show promising results in a traffic network simulated in the simulation of urban mobility traffic simulator, without suffering from instability issues during the training process.
    URI
    http://hdl.handle.net/10379/12999
    Collections
    • Externally hosted open access publications with NUI Galway authors (2)

    Related items

    Showing items related by title, author, creator and subject.

    • Parallel reinforcement learning for traffic signal control 

      Mannion, Patrick; Duggan, Jim; Howley, Enda (Elsevier BV, 2015-01-01)
    • Cloud computing and consumer electronics: a perfect match or a hidden storm? [Soapbox] 

      Corcoran, Peter M. (IEEE, 2012-04)
      Cloud computing has emerged strongly into the consumer domain. As industry competes to gain market share from the rapidly growing multitudes of cloud-IT users, both network and data center infrastructures will grow more ...
    • Influence agenda setting through Twitter for health promotion 

      Albalawi, Yousef (2016-05-12)
      We are in what is known as the new media era, and it impacts all dimensions and aspects of people’s lives. Through advanced technology and the internet, new media continues evolving to change people’s lives so that they ...
    • Copyright @ NUI Galway
    • Library
    • NUI Galway