Skip to content
MIT Press
  • MIT Press
  • Books
    • Column
      • View all subjects
      • New releases
      • Catalogs
      • Textbooks
      • Series
      • Awards
    • Column
      • Authors
      • Distributed presses
      • The MIT Press Reader
      • Podcasts
      • Collections
    • Column
      • MIT Press Direct

        MIT Press Direct is a distinctive collection of influential MIT Press books curated for scholars and libraries worldwide.

        • Learn more
  • Journals
    • column
      • Journals all topics
      • Economics
      • International Affairs, History, & Political Science
    • column
      • Arts & Humanities
      • Science & Technology
      • Open access
    • column
      • MIT Press journals

        MIT Press began publishing journals in 1970 with the first volumes of Linguistic Inquiry and the Journal of Interdisciplinary History. Today we publish over 30 titles in the arts and humanities, social sciences, and science and technology.

        • Learn more
  • Open Access
    • column
      • Open access at the MIT Press
      • Open access books
      • Open access journals
    • column
      • Direct to Open
      • MIT Open Publishing Services
      • MIT Press Open on PubPub
    • Column
      • Open access

        The MIT Press has been a leader in open access book publishing for over two decades, beginning in 1995 with the publication of William Mitchell’s City of Bits, which appeared simultaneously in print and in a dynamic, open web edition.

        • Learn more
  • Resources
    • column
      • Current authors
      • Prospective authors
      • Instructors
    • column
      • Media inquiries
      • Booksellers
      • Rights and permissions
    • column
      • Resources

        Collaborating with authors, instructors, booksellers, librarians, and the media is at the heart of what we do as a scholarly publisher. If you can’t find the resource you need here, visit our contact page to get in touch.

        • Learn more
  • Give
  • About
    • Column
      • About
      • Jobs
      • Internships
      • MIT Press Editorial Board
      • MIT Press Management Board
      • Our MIT story
    • Column
      • Catalogs
      • News
      • Events
      • Conferences
      • Bookstore
    • Column
      • The MIT Press

        Established in 1962, the MIT Press is one of the largest and most distinguished university presses in the world and a leading publisher of books and journals at the intersection of science, technology, art, social science, and design.

        • Learn more
  • Contact Us
Newsletter
MIT Press
Newsletter

Books

    Authors

      On the site

        • Home
        • Adaptive Computation and Machine Learning series
        • computers
        • mathematics
        • Distributional Reinforcement Learning
        Distributional Reinforcement Learning

        Adaptive Computation and Machine Learning series

        Distributional Reinforcement Learning

        by Marc G. Bellemare, Will Dabney and Mark Rowland

        • $60.00 Hardcover

        384 pp., 6 x 9 in, 48 figures

        • Hardcover
        • 9780262048019
        • Published: May 30, 2023
        • Publisher: The MIT Press

        $60.00

        • Penguin Random House
        • Amazon
        • Barnes and Noble
        • Bookshop.org
        • Indiebound
        • Indigo
        • Books a Million

        Other Retailers:

        • Amazon.co.uk
        • Blackwells
        • Bookshop.org
        • Foyles
        • Hive
        • Waterstones
        • Penguin Random House
        • Amazon
        • Barnes and Noble
        • Bookshop.org
        • Indiebound
        • Indigo
        • Books a Million
        • Request permissions
        • Description
        • Author(s)
        • Praise

        The first comprehensive guide to distributional reinforcement learning, providing a new mathematical formalism for thinking about decisions from a probabilistic perspective.

        Distributional reinforcement learning is a new mathematical formalism for thinking about decisions. Going beyond the common approach to reinforcement learning and expected values, it focuses on the total reward or return obtained as a consequence of an agent's choices—specifically, how this return behaves from a probabilistic perspective. In this first comprehensive guide to distributional reinforcement learning, Marc G. Bellemare, Will Dabney, and Mark Rowland, who spearheaded development of the field, present its key concepts and review some of its many applications. They demonstrate its power to account for many complex, interesting phenomena that arise from interactions with one's environment.

        The authors present core ideas from classical reinforcement learning to contextualize distributional topics and include mathematical proofs pertaining to major results discussed in the text. They guide the reader through a series of algorithmic and mathematical developments that, in turn, characterize, compute, estimate, and make decisions on the basis of the random return. Practitioners in disciplines as diverse as finance (risk management), computational neuroscience, computational psychiatry, psychology, macroeconomics, and robotics are already using distributional reinforcement learning, paving the way for its expanding applications in mathematical finance, engineering, and the life sciences. More than a mathematical approach, distributional reinforcement learning represents a new perspective on how intelligent agents make predictions and decisions.

        Marc G. Bellemare is Senior Staff Research Scientist, Google Research and Adjunct Professor, McGill University.

        Will Dabney is Senior Staff Research Scientist, DeepMind.

        Mark Rowland is Senior Research Scientist, DeepMind.

        “This important book extends the machinery of reinforcement learning to encompass full outcome statistics. The brain inevitably got there first, but the authors' brilliant insights show this, how, and why.”

        Peter Dayan, Director, Max Planck Institute for Biological Cybernetics, Tübingen

        “This book shows how reinforcement learning agents can estimate and use full reward distributions to make better decisions. It is a significant enhancement that opens a new range of applications.”

        Andrew Barto, Professor Emeritus, University of Massachusetts Amherst; coauthor of Reinforcement Learning: An Introduction, second edition (MIT Press)

        Related Books

        The Developmental Organization of Robot Behavior
        Neurorobotics
        Microprediction
        Awkward Intelligence
        Working with AI
        Tales from a Robotic World
        How to Stay Smart in a Smart World
        Machine Learning from Weak Supervision
        Artificial Communication
        Machines like Us
        logo
        • Column 1
          • Books
          • Journals
          • The MIT Press Reader
          • Podcasts
          • Imprints
        • Column 2
          • The MIT Press
            • About
            • Bookstore
            • Catalogs
            • Conferences
            • Press Editorial Board
            • Jobs
            • Internships
            • Press Management Board
            • News
            • Staff
            • Code of Conduct
            • Give
        • Column 3
          • Site Help
            • Accessibility
            • FAQ
            • Our eBooks
            • Privacy Policy
            • Terms of Use
        • Column 4
          • Resources
            • Current Authors
            • Prospective Authors
            • Booksellers
            • Instructors
            • Rights and Permissions
            • Media Inquiries
            • MIT Discounts
        • Column 5
          • Digital
            • CogNet
            • Digital Partners and Products
            • Knowledge Futures Group
            • MIT Press Direct
        • US

          One Broadway 12th Floor Cambridge, MA 02142

        • UK

          Unit 57710 PO Box 6945 London W1A 6US UK

        • Contact

        Connect

        © 2023 MIT Press. All Rights Reserved.

        Powered by Supadu