IDEAS home Printed from https://ideas.repec.org/a/plo/pone00/0230432.html
   My bibliography  Save this article

ToyArchitecture: Unsupervised learning of interpretable models of the environment

Author

Listed:
  • Jaroslav Vítků
  • Petr Dluhoš
  • Joseph Davidson
  • Matěj Nikl
  • Simon Andersson
  • Přemysl Paška
  • Jan Šinkora
  • Petr Hlubuček
  • Martin Stránský
  • Martin Hyben
  • Martin Poliak
  • Jan Feyereisl
  • Marek Rosa

Abstract

Research in Artificial Intelligence (AI) has focused mostly on two extremes: either on small improvements in narrow AI domains, or on universal theoretical frameworks which are often uncomputable, or lack practical implementations. In this paper we attempt to follow a big picture view while also providing a particular theory and its implementation to present a novel, purposely simple, and interpretable hierarchical architecture. This architecture incorporates the unsupervised learning of a model of the environment, learning the influence of one’s own actions, model-based reinforcement learning, hierarchical planning, and symbolic/sub-symbolic integration in general. The learned model is stored in the form of hierarchical representations which are increasingly more abstract, but can retain details when needed. We demonstrate the universality of the architecture by testing it on a series of diverse environments ranging from audio/visual compression to discrete and continuous action spaces, to learning disentangled representations.

Suggested Citation

  • Jaroslav Vítků & Petr Dluhoš & Joseph Davidson & Matěj Nikl & Simon Andersson & Přemysl Paška & Jan Šinkora & Petr Hlubuček & Martin Stránský & Martin Hyben & Martin Poliak & Jan Feyereisl & Marek Ros, 2020. "ToyArchitecture: Unsupervised learning of interpretable models of the environment," PLOS ONE, Public Library of Science, vol. 15(5), pages 1-50, May.
  • Handle: RePEc:plo:pone00:0230432
    DOI: 10.1371/journal.pone.0230432
    as

    Download full text from publisher

    File URL: https://journals.plos.org/plosone/article?id=10.1371/journal.pone.0230432
    Download Restriction: no

    File URL: https://journals.plos.org/plosone/article/file?id=10.1371/journal.pone.0230432&type=printable
    Download Restriction: no

    File URL: https://libkey.io/10.1371/journal.pone.0230432?utm_source=ideas
    LibKey link: if access is restricted and if your library uses this service, LibKey will redirect you to where you can use your library subscription to access this item
    ---><---

    References listed on IDEAS

    as
    1. Karl Friston, 2008. "Hierarchical Models in the Brain," PLOS Computational Biology, Public Library of Science, vol. 4(11), pages 1-24, November.
    2. Karl J Friston & Jean Daunizeau & Stefan J Kiebel, 2009. "Reinforcement Learning or Active Inference?," PLOS ONE, Public Library of Science, vol. 4(7), pages 1-13, July.
    3. Nicolas Brodu, 2011. "Reconstruction Of Epsilon-Machines In Predictive Frameworks And Decisional States," Advances in Complex Systems (ACS), World Scientific Publishing Co. Pte. Ltd., vol. 14(05), pages 761-794.
    4. Volodymyr Mnih & Koray Kavukcuoglu & David Silver & Andrei A. Rusu & Joel Veness & Marc G. Bellemare & Alex Graves & Martin Riedmiller & Andreas K. Fidjeland & Georg Ostrovski & Stig Petersen & Charle, 2015. "Human-level control through deep reinforcement learning," Nature, Nature, vol. 518(7540), pages 529-533, February.
    Full references (including those not matched with items on IDEAS)

    Most related items

    These are the items that most often cite the same works as this one and are cited by the same works as this one.
    1. Mateus Joffily & Giorgio Coricelli, 2013. "Emotional Valence and the Free-Energy Principle," Post-Print halshs-00834063, HAL.
    2. Dongqi Han & Kenji Doya & Dongsheng Li & Jun Tani, 2024. "Synergizing habits and goals with variational Bayes," Nature Communications, Nature, vol. 15(1), pages 1-14, December.
    3. Tulika Saha & Sriparna Saha & Pushpak Bhattacharyya, 2020. "Towards sentiment aided dialogue policy learning for multi-intent conversations using hierarchical reinforcement learning," PLOS ONE, Public Library of Science, vol. 15(7), pages 1-28, July.
    4. Mahmoud Mahfouz & Angelos Filos & Cyrine Chtourou & Joshua Lockhart & Samuel Assefa & Manuela Veloso & Danilo Mandic & Tucker Balch, 2019. "On the Importance of Opponent Modeling in Auction Markets," Papers 1911.12816, arXiv.org.
    5. Imen Azzouz & Wiem Fekih Hassen, 2023. "Optimization of Electric Vehicles Charging Scheduling Based on Deep Reinforcement Learning: A Decentralized Approach," Energies, MDPI, vol. 16(24), pages 1-18, December.
    6. Jacob W. Crandall & Mayada Oudah & Tennom & Fatimah Ishowo-Oloko & Sherief Abdallah & Jean-François Bonnefon & Manuel Cebrian & Azim Shariff & Michael A. Goodrich & Iyad Rahwan, 2018. "Cooperating with machines," Nature Communications, Nature, vol. 9(1), pages 1-12, December.
      • Abdallah, Sherief & Bonnefon, Jean-François & Cebrian, Manuel & Crandall, Jacob W. & Ishowo-Oloko, Fatimah & Oudah, Mayada & Rahwan, Iyad & Shariff, Azim & Tennom,, 2017. "Cooperating with Machines," TSE Working Papers 17-806, Toulouse School of Economics (TSE).
      • Abdallah, Sherief & Bonnefon, Jean-François & Cebrian, Manuel & Crandall, Jacob W. & Ishowo-Oloko, Fatimah & Oudah, Mayada & Rahwan, Iyad & Shariff, Azim & Tennom,, 2017. "Cooperating with Machines," IAST Working Papers 17-68, Institute for Advanced Study in Toulouse (IAST).
      • Jacob Crandall & Mayada Oudah & Fatimah Ishowo-Oloko Tennom & Fatimah Ishowo-Oloko & Sherief Abdallah & Jean-François Bonnefon & Manuel Cebrian & Azim Shariff & Michael Goodrich & Iyad Rahwan, 2018. "Cooperating with machines," Post-Print hal-01897802, HAL.
    7. Sun, Alexander Y., 2020. "Optimal carbon storage reservoir management through deep reinforcement learning," Applied Energy, Elsevier, vol. 278(C).
    8. Yassine Chemingui & Adel Gastli & Omar Ellabban, 2020. "Reinforcement Learning-Based School Energy Management System," Energies, MDPI, vol. 13(23), pages 1-21, December.
    9. Woo Jae Byun & Bumkyu Choi & Seongmin Kim & Joohyun Jo, 2023. "Practical Application of Deep Reinforcement Learning to Optimal Trade Execution," FinTech, MDPI, vol. 2(3), pages 1-16, June.
    10. Lu, Yu & Xiang, Yue & Huang, Yuan & Yu, Bin & Weng, Liguo & Liu, Junyong, 2023. "Deep reinforcement learning based optimal scheduling of active distribution system considering distributed generation, energy storage and flexible load," Energy, Elsevier, vol. 271(C).
    11. Yuhong Wang & Lei Chen & Hong Zhou & Xu Zhou & Zongsheng Zheng & Qi Zeng & Li Jiang & Liang Lu, 2021. "Flexible Transmission Network Expansion Planning Based on DQN Algorithm," Energies, MDPI, vol. 14(7), pages 1-21, April.
    12. Huang, Ruchen & He, Hongwen & Gao, Miaojue, 2023. "Training-efficient and cost-optimal energy management for fuel cell hybrid electric bus based on a novel distributed deep reinforcement learning framework," Applied Energy, Elsevier, vol. 346(C).
    13. Michelle M. LaMar, 2018. "Markov Decision Process Measurement Model," Psychometrika, Springer;The Psychometric Society, vol. 83(1), pages 67-88, March.
    14. Micha Heilbron & Florent Meyniel, 2019. "Confidence resets reveal hierarchical adaptive learning in humans," PLOS Computational Biology, Public Library of Science, vol. 15(4), pages 1-24, April.
    15. Zichen Lu & Ying Yan, 2024. "Temperature Control of Fuel Cell Based on PEI-DDPG," Energies, MDPI, vol. 17(7), pages 1-19, April.
    16. Yang, Ting & Zhao, Liyuan & Li, Wei & Zomaya, Albert Y., 2021. "Dynamic energy dispatch strategy for integrated energy system based on improved deep reinforcement learning," Energy, Elsevier, vol. 235(C).
    17. Wang, Xuan & Shu, Gequn & Tian, Hua & Wang, Rui & Cai, Jinwen, 2020. "Operation performance comparison of CCHP systems with cascade waste heat recovery systems by simulation and operation optimisation," Energy, Elsevier, vol. 206(C).
    18. Wang, Yi & Qiu, Dawei & Sun, Mingyang & Strbac, Goran & Gao, Zhiwei, 2023. "Secure energy management of multi-energy microgrid: A physical-informed safe reinforcement learning approach," Applied Energy, Elsevier, vol. 335(C).
    19. Parvez Farazi, Nahid & Zou, Bo & Tulabandhula, Theja, 2022. "Dynamic On-Demand Crowdshipping Using Constrained and Heuristics-Embedded Double Dueling Deep Q-Network," Transportation Research Part E: Logistics and Transportation Review, Elsevier, vol. 166(C).
    20. Louback, Eduardo & Biswas, Atriya & Machado, Fabricio & Emadi, Ali, 2024. "A review of the design process of energy management systems for dual-motor battery electric vehicles," Renewable and Sustainable Energy Reviews, Elsevier, vol. 193(C).

    More about this item

    Statistics

    Access and download statistics

    Corrections

    All material on this site has been provided by the respective publishers and authors. You can help correct errors and omissions. When requesting a correction, please mention this item's handle: RePEc:plo:pone00:0230432. See general information about how to correct material in RePEc.

    If you have authored this item and are not yet registered with RePEc, we encourage you to do it here. This allows to link your profile to this item. It also allows you to accept potential citations to this item that we are uncertain about.

    If CitEc recognized a bibliographic reference but did not link an item in RePEc to it, you can help with this form .

    If you know of missing items citing this one, you can help us creating those links by adding the relevant references in the same way as above, for each refering item. If you are a registered author of this item, you may also want to check the "citations" tab in your RePEc Author Service profile, as there may be some citations waiting for confirmation.

    For technical questions regarding this item, or to correct its authors, title, abstract, bibliographic or download information, contact: plosone (email available below). General contact details of provider: https://journals.plos.org/plosone/ .

    Please note that corrections may take a couple of weeks to filter through the various RePEc services.

    IDEAS is a RePEc service. RePEc uses bibliographic data supplied by the respective publishers.