FY2019 Annual Report

Neural Computation Unit
Professor Kenji Doya

Group photo

 

Abstract

The Neural Computation Unit pursues the dual goals of developing robust and flexible learning algorithms and elucidating the brain’s mechanisms for robust and flexible learning. Our specific focus is on how the brain realizes reinforcement learning, in which an agent, biological or artificial, learns novel behaviors in uncertain environments by exploration and reward feedback. We combine top-down, computational approaches and bottom-up, neurobiological approaches to achieve these goals.

In FY2019, we continued research along three major externally funded projects: Kakenhi project on Artificial Intelligence and Brain Science, Brain/MINDS project, and Post-K Supercomputing project on Brain and Artificial Intelligence.

In the Kakenhi project on Artificial Intelligence and Brain Science, we published perspective papers on what we can learn from the brain for next-generation AI. Our PhD students, Paavo and Tadashi, gave talks on reinforcement learning algorithms at AISTATS and NeurIPS Workshop, and completed their theses after enjoying internships at DeepMind Paris.

In Japan's major brain science project, Brain/MINDS, we submitted our paper on optimization and validation of diffusion-MRI-based fiber tracking algorithms using neural tracer data as the reference. We constructed a whole-brain mesoscopic networks mode using the connecteme derived by the optimized algortihm and showed that the model could approximate the functional connectivity from functional MRI data.

In the Post-K supercomputing project on Brain and Artificial Intellignece, we built biologically constrained spiking neural network model of the basal ganglia circuit and integrated that with the cerebellum and thalamo-cortical network models constructed by our collaboratos at UEC and RIKEN for whole-brain simulations.

1. Staff

Systems Neurobiology Group

  • Katsuhiko Miyazaki, Staff Scientist
  • Kayoko Miyazaki, JSPS Research Fellow
  • Kazumi Kasahara, JSPS Research Fellow
  • Yukako Yamane, JSPS Research Fellow
  • Yuzhe Li, Postdoctoral Scholar
  • Sergey Zobnin, OIST Student
  • Masakazu Taira, OIST Student
  • Miles Desforges, OIST Student
  • Tomohiko Yoshizawa, Visiting Researcher (TMDU)

Dynamical Systems Group

  • Carlos Enrique Gutierrez, Postdoctoral Scholar
  • Hiromichi Tsukada, Postdoctoral Scholar
  • Alessandro La Chioma, Postdoctoral Scholar
  • Jessica Verena Schulze, Junior Research Fellow
  • Junichiro Yoshimoto, Visiting Researcher (NAIST)
  • Tomoki Tokuda, Visiting Researcher (NAIST, ATR)
  • Hiroaki Hamada, Visiting Researcher (ARAYA)

Adaptive Systems Group

  • Christopher Buckley, Technician
  • Shoko Igarashi, OIST Student
  • Farzana Rahmen, OIST Student
  • Tadashi Kozuno, OIST Student/Junior Research Fellow
  • Paavo Parmas, OIST Student/Junior Research Fellow
  • Ho Ching Chiu, OIST Student
  • Kristine Faith Roque, OIST Student

Research Unit Administrators

  • Emiko Asato
  • Kikuko Matsuo
  • Misuzu Saito

2. Collaborations

  • Dr. Remi Munos at DeepMind Paris and Dr. Masashi Sugiyama at RIKEN AIP on reinforcement learning algorithms.
  • Prof. Shin Ishii lab at Kyoto University and researchers at RIKEN Center for Brain Science for neural data analysis in the Brain/MINDS Project.
  • Prof. Benoit Girard at Sorbonne University, Dr. Marcus Diesmann at Julich Research Center, Dr. Jun Igarashi at RIKEN, and Prof. Tadashi Yamazaki at University of Electro-Communications for building whole-brain spiking networks models in Post-K Supercomputing Project

3. Activities and Findings

3.1 Nuerobiology Experiments[Systems Neurobilogy Group]

3.1.1 Neural substrate of dynamic Bayesian inference [Kakenhi Project on AI and Brain Science]

We continued our experiments to clarify how predictive information from actions and sensory information through environmental interaction are integrated across different layers of the  somatosensory and motor cortical circuits in mice. We trained mice to perform active level manipulation tasks and started simultaneous cross-layer calcium imaging by a prism lens.

3.1.2 The role of serotonin in the regulation of patience [Kakenhi Project on AI and Brain Science] 

In order to clarify the role of serotonin in behaviors, we performed neural recording, microdialysis measurement, and optogenetic manipulation of serotonin neural activity from the dorsal raphe nucleus (DRN), the major source of serotonergic projection to the cortex and the basal ganglia (Miyazaki et al., 2016, 2018). We further performed optogenetic stimulation of serotonergic axon terminals in the orbitofrontal cortex (OFC), medial prefrontal cortex (mPFC) and the nucleus accumbens and found differential effects on promoting waiting for delayed rewards (Miyazaki et al. SfN 2019). We extended our Bayesian decision model (Miyazaki et al. 2018) to reproduced the new experimental results.

3.2. Neural Data Analysis and Modeling [Dynamical Systems Group]

3.2.1 Analysis and modeling of marmoset brain data [Brain/MINDS Project]

We develop data-analysis and modeling methods to utilize the variety of marmoset brain data obtained in the Brain/MIDNS program. We developed a framework for optimizing and validating diffusion MRI-based fiber-tracking algorithms in reference to neural tracer data and showed that long-range connections are more accurately identified by the optimized algorithm (Gutierrez et al. arXiv:1911.13215).

Using the strutural connectivity derived by the optimized algorithm, we constructed a whole cortical network model with 60 regions of interest (ROI), where each ROI is approximated a Wilson-Cowan model. By adjusting the excitatory and inhibitory connections, we could reproduce the functional connectivity derived from resting-state functional MRI data (Tsukada et al., JSMBE 2019).

3.2.2 Large-scale spiking network models of the basal ganglia [Post-K Supercomputing Project on Brain and Artificial Intelligence]

We constructed an integrate-and-fire neural network model of the basal ganglia circuit based on the anatomical and physiological data and explored the mechanisms for action selection. We further connected the basal ganglia model with the cerebellar and thalamo-cortical network models constructed by our collaborators at UEC and RIKEN (Gutierrez et al., CNS*2019).

3.3 Robotics and Reinforcement Learning [Adaptive Systems Group]

3.3.1 Data-efficient reinforcement learning [Kakenhi Project on AI and Brain Science]

While the combination of reinforcement learning with deep neural networks has shown remarkable performance, the requirement of huge data for training is a major issue in robotic applications. We formulated "Conservative Value Iteration" algorithm and thoeretically proved its data efficiency under sampling noise and approximation errors (Kozuno et al., AISTATS 2019).

We derived an efficient gradient computation method for general probabilistic graphic models and clarified the relationship between previous methods (Parmas and Sugiyama, NeurIPS 2019 Deep RL Workshop).

3.3.2 Development of smartphone robot platform [Kakenhi Project on AI and Brain Science]

We updated our smartphone robot platform to make its electronics more stable and software compatible with the latest Android OSs. The robots are being utilized for a variety of research topics, such as benchmarks for data-efficient reinforcement learning (Chiu & Doya, Brain and Mind Workshop 2020), evolution of alternative reproductive strategies (Rahman et al., AROB2019), and learning to cooperate with symbolic communication.

Figure 1: The smartphone robot (2019  version).

4. Publications

4.1 Journals

  1. Doya K, Taniguchi T (2019). Toward evolutionary and developmental intelligence. Current Opinion in Behavioral Sciences, 29, 91-96. https://doi.org/10.1016/j.cobeha.2019.04.006
  2. Kozuno T, Uchibe E, Doya K (2019). Theoretical analysis of efficiency and robustness of softmax and gap-increasing operators in reinforcement learning. Proceedings of Machine Learning Research (PMLR), 89, 2995-3003. http://proceedings.mlr.press/v89/kozuno19a.html
  3. Doya K (2019). Appreciation for the JNNS Academic Award. The Brain and Neural Networks, 26, 159-164 [銅谷賢治. 日本神経回路学会学術賞にあたって. 日本神経回路学会誌]. https://doi.org/10.3902/jnns.26.159
  4. Doya K, Matsuo Y (2019). Artificial intelligence and brain science: The present and the future. BRAIN and NERVE, 71, 649-655 [銅谷賢治, 松尾豊 (2019). 人工知能と脳科学の現在とこれから. BRAIN and NERVE]. https://doi.org/10.11477/mf.1416201337

4.2 Books and other one-time publications

  1. Moren J, Igarashi J, Shouno O, Yoshimoto J, Doya K (2019). Dynamics of basal ganglia and thalamus in Parkinsonian tremor. in Cutsuridis V (ed). Multiscale Models of Brain Disorders, Springer. http://doi.org/10.1007/978-3-030-18830-6
  2. Doya K (2019). What will adaptive autonomous robots dream of? in Artificial Intelligence Art and Aesthetics Research Group (ed.), Artificial Intelligence Art and Aesthetics Exhibition - Archive Collection, 118-119 [銅谷賢治. 自律学習ロボットは何の夢を見るか. 人工知能美学芸術研究会編, 人工知能美学芸術展記録集]. https://www.aibigeiken.com/store/aiaae_ac_e.html
  3. Doya K (2019). Looking back at artificial intelligence art and aesthetics exhibition in OIST. in Artificial Intelligence Art and Aesthetics Research Group (ed.), Artificial Intelligence Art and Aesthetics Exhibition - Archive Collection,170-171 [銅谷賢治. 人工知能美学芸術展 in OISTを振り返って. 人工知能美学芸術展記録集, 人工知能美学芸術研究会]. https://www.aibigeiken.com/store/aiaae_ac_e.html
  4. Parmas P (2020). Total stochastic gradient algorithms and applications to model-based reinforcement learning. PhD Thesis, Okinawa Institute of Science and Technology Graduate University. http://doi.org/10.15102/1394.00001064
  5. Kozuno T (2020). Efficient and noise-tolerant reinforcement learning algorithms via theoretical analysis of gap-Increasing and softmax operators. PhD Thesis, Okinawa Institute of Science and Technology Graduate University. http://doi.org/10.15102/1394.00001389

4.3 Oral and Poster Presentations

  1. Chiu HC, Doya K (2020). Representation and grounding of abstract concepts: a preliminary investigation. 20th Winter Workshop on the Mechanism of Brain and Mind: Brain and Artificial Intelligence. Rusutsu, Hokkaido.
  2. Doya K (2019). Neural circuits for mental simulation. Keynote Lecture, 28th Annual Computational Neuroscience Meeting (CNS*2019). Universitat de Barcelona, Barcelona, Spain.
  3. Doya K (2019). Possible roles of dopamine in model-free and model-based decision and learning. Dopaminergic Signaling Workshop, 28th Annual Computational Neuroscience Meeting (CNS*2019). Universitat de Barcelona, Barcelona, Spain.
  4. Doya K (2019). Ethologically grounded motivation and neural implementation of mental simulation. Fourth International Workshop on Intrinsically-Motivated Open-ended Learning (IMOL2019). Institute for Advanced Studies (FIAS) Frankfurt, Germany.
  5. Doya K (2019). What can we further learn from the brain for AI and robotics? Global AI Summit, Seoul National University, Korea.
  6. Doya K (2019). Neural circuits for mental simulation. International Conference on Cognitive Science (ICCS) 2019. Seoul National University, Korea.
  7. Doya K (2019). Artificial intelligence and brain science. The Joint Symposium of WPI-IIIS, Ph.D. Program in Humanics, and 36th Takamine conference. Tokyo Conference Center Shinagawa, Tokyo.
  8. Doya K (2019). Toward whole-brain multi-scale modeling. Neuroinformatics and Neurobiology (NBNI2019). Chongqing, China.
  9. Doya K (2019). Learning to communicate for ecological fitness. Shonan Meeting No. 141. Language as Goal-Directed Sequential Behavior: Computational Theories, Brain Mechanisms, Evolutionary Roots. Shonan Village Center, Kanagawa.
  10. Doya K (2019). Big data challenges in neuroscience. IEEE CIS Summer School 2019 "Big Data Analytics and Stream Processing: Tools, Techniques and Application". Indian Institute of Information Technology Allahabad, India.
  11. Doya K (2019). Neural circuits for reinforcement learning and mental simulation. IBRO 2019 Symposium: Valence and Reward Encoding. Deagu EXCO, South Korea.
  12. Doya K (2019). Reinforcement learning in machines and the brain. Keynote Lecture, Conference on Robot Learning (CoRL 2019). Senri Life Science Center, Osaka.
  13. Doya K (2019). Systems biology of reinforcement learning. International Conference on Systems Biology (ICSB 2019). OIST Auditorium, Okinawa.
  14. Doya K (2019). What can we further learn from the brain for cognitive robotics? IROS 2019 Workshop: Deep Probabilistic Generative Models for Cognitive Architecture in Robotics. The Venetian Macao, Macau.  
  15. Doya K (2019). Patience, confidence and serotonin. Blue Brain Seminar. EPFL Geneva, Switzerland.
  16. Doya k (2019). Consciousness as data assimilation. ARAYA Consciousness Club. Roppongi, Tokyo.
  17. Doya K (2019). What can we learn from the brain for next ai? The 3rd Ryudai-OIST Symposium: Basic Medical Science to Clinical Medicine. Sydney Brenner Lecture Theatre, OIST, Okinawa.  
  18. Doya K (2019). Bayesian brain today. 36th Annual Meeting of Japanese Cognitive Science Society, Hamamatsu.  [銅谷賢治. ベイジアンブレインの今日. 第36回日本認知科学学会大会, 静岡大学浜松キャンパス].
  19. Doya K (2019). Mental simulation in mice and evolving rewards in robots. 12th Human Behavior and Evolution Society Japan. Meiji Gakuin University, Tokyo.
  20. Doya K (2019). Artificial intelligence and brain science. 58th Annual Meeting of Japanese Biomedical Engineering Society. Okinawa Convention Center [銅谷賢治. 人工知能と脳科学. 第58回日本生体医工学会大会. 沖縄コンベンションセンター].
  21. Doya K (2020). What can we further learn from the brain for artificial intelligence? 20th Winter Workshop on the Mechanism of Brain and Mind. Rusutsu, Hokkaido.
  22. Doya K (2020). How can the brain connect predictors and actors on the fly? Workshop on Learning for flexible, context-sensitive behavior. Forschungssekretariat, Kunst am ZiF, Germany.
  23. Doya K (2020). Computation and neural implementation of reinforcement learning. Seminar at LNC2. ENS Paris, France.
  24. Gutierrez CE (2019). Building a whole-brain model. 3rd Ryudai-OIST symposium. OIST, Okinawa.
  25. Gutierrez CE (2019). Toward realizing reinforcement learning in a whole-brain spiking neural network model. 20th Winter Workshop on the Mechanism of Brain and Mind. Rusutsu, Hokkaido.
  26. Gutierrez CE (2019). Large-scale simulation of a spiking neural network model consisting of cortex, thalamus, cerebellum and basal ganglia on K computer. NEST conference. NMBU, As, Norway.   
  27. Gutierrez CE, Igarashi J, Sun Z, Lienard J, Yamaura H, Yamazaki T, Morteza H, Girard B, Arbuthnott G, Plesser H, Diesmann M, Doya K (2019). A spiking neural network model of the whole-brain circuit linking basal ganglia, cerebellum and cortex. Neuro2019. Toki Messe, Niigata.
  28. Gutierrez CE, Igarashi J, Sun Z, Yamaura H, Yamazaki T, diesmann M, Lienard J, Morteza H, Girard B, Arbuthnott G, Plesser EH, Doya K (2019). A whole-brain spiking neural network model linking basal ganglia, cerebellum, cortex and thalamus. CNS*2019. Barcelona, Spain.
  29. Kasahara K (2019). Individual variability of brain-machine interface. The 3rd Ryudai-OIST symposium. OIST, Okinawa.
  30. Kasahara K, DaSalla C, Honda M, Hanakawa T (2019). Individual variability of brain-machine interface. 58th Annual Conference of Japanese Society for Medical and Biological Engineering. Okinawa Convention Center, Ginowan-city, Okinawa.
  31. Kasahara K, Doya K (2019). The enhancement of functional connectivity induced by longitudinal motor training in mice. 58th Annual Conference of Japanese Society for Medical and Biological Engineering. Okinawa Convention Center, Ginowan-city, Okinawa.
  32. Kozuno T, Uchibe E, Doya K (2019). Theoretical analysis of efficiency and robustness of softmax and gap-increasing operators in reinforcement learning. The 22nd International Conference on Artificial Intelligence and Statistics (AISTATS 2019). Naha, Okinawa.
  33. Han D, Doya K, Tani J (2019). Self-organization of action hierarchy and compositionality by reinforcement learning with recurrent networks. NeurIPS 2019 Workshop on Context and Compositionality in Biological and Artificial Neural Systems. Vancouver, Canada. (arXiv:1901.10113)
  34. Han D, Doya K, Tani J (2020). Variational recurrent models for solving partially observable control tasks. International Conference on Learning Representations (ICLR 2020). Online. (arXiv:1912.10703)
  35. Miyazaki K (2019). Serotonergic control mechanism of reward waiting behavior. 254th Tsukuba Brain Science Seminar, University of Tsukuba. [宮崎勝彦, セロトニンによる報酬待機行動の制御機構. 第254回筑波脳科学セミナー, 筑波大学].
  36. Miyazaki KW, Miyazaki K, Yamanaka A, Tanaka KF, Doya K (2019). Stimulation of serotonergic terminals in the orbitofrontal and medial prefrontal cortices differentially affects waiting for the future rewards. Society for Neuroscience 2019. Chicago, USA.
  37. Nara S, Tsukada H, Fujii H, Tsuda I (2019). A three-modules scenario in an interpretation of visual hallucination in dementia with lewy bodies and preliminary results of computer experiments. IJCNN 2019. 2019 International Joint Conference on Neural Networks (IJCNN). Budapest, Hungary.
  38. Ota S, Doya K (2019). Intrinsic motivation in creative activity: A human behavioral experiment for identifying the factors that influence intrinsic motivation. WIML, NeurIPS. Vancouver, Canada.
  39. Ota S, Doya K (2019). Intrinsic motivation in play: Preliminary experiment for analyzing how learning environmental condition influences intrinsic motivation. International Conference on Cognitive Science 2019. Seoul National University, Korea.
  40. Parmas P, Sugiyama M (2019). A unified view of likelihood ratio and reparameterization gradients and an optimal importance sampling scheme. Deep Reinforcement Learning Workshop, NeurIPS. Vancouver, Canada.
  41. Rahman F, Doya K, Mickheyev A (2020). Identifying the evolutionary conditions for the emergence of alternative reproductive tactics in simulated robot colonies. AROB 25th 2020. B-con Plaza, Oita.   
  42. Tokuda T (2019). Multiple co-clustering with heterogenous marginal distributions and its application to identify subtypes of depressive disorder. 62nd ISI World Statistics Congress 2019. Kuala Lumpur, Malaysia.
  43. Tsukada H (2019). MRI data-driven model of the whole marmoset brain and local neural circuit model of memory. National Institute of Physiological Science Workshop: Understanding the brain and neurla circuits from dynamical systems viewpoint. NIPS, Okazaki. [MRIデータ駆動型マーモセット全脳モデルと局所神経回路記憶モデル. 生理学研究所、愛知, 生理研研究会: 力学系の視点からの脳・神経回路の理解].
  44. Tsukada H, Ane L, Hata J, Hamada H, Nakae K, Gutierrez C, Skibbe H, Woodward A, Ishii S, Okano H, Gustavo D, Doya K (2019). Development of data-driven integrate model using Brain/MINDS database. Brain/MINDS kickoff meeting. Hakone, Kanagawa.
  45. Tsukada H, Lopez A, Hata J, Hamada H, Nakae K, Gutierrez C, Skibbe H, Woodward A, Ishii S, Okano H, Gustavo D, Doya K (2019). Whole brain modeling and dynamic analysis using structural and functional MRI data of marmosets. 58th Annual Meeting of Japanese Biomedical Engineering Society. Okinawa Convention Center, Ginowan, Okinawa.  
  46. Tsukada H, Tsukada M (2019). Context-dependent learning and memory based on spatio-temporal learning rule. 7th International Conference on Cognitive Neurodynamics (ICCN2019). Alghero, Italy.
  47. Tsukada M, Tsukada H (2019). Fractal structure in hokusai's "great wave" and the memory neural network - brain, art and ai -.7th International Conference on Cognitive Neurodynamics (ICCN2019). Alghero, Italy.

5. Intellectual Property Rights and Other Specific Achievements

Under OIST Proof of Concept Program, Paavo Parmas started to implement his Total Stochastic Gradient Method (US Patent Application 62/749,908) as an open-source library to promote industrial applications.

6. Meetings and Events

6.1 Seminars

Bayesian Inference and Experimental Design for Implicit Models

  • Date:  April 19, 2019
  • Venue: OIST Campus Lab1
  • Speaker: Dr Michael Gutmann (School of Informatics, University of Edinburgh)

On the way to the automated discovery of novel patterns and dynamics in physical and chemical systems by intrinsically motivated deep learning

  • Date: May 30, 2019
  • Venue: OIST Campus Lab1
  • Speaker: Chris Reinke (Flowers Lab, Inria Bordeaux, France)

Effects of a professional classical solo singer education on vocal tract adjustments during singing - A Longitudinal Study

  • Date:  June 11, 2019
  • Venue: OIST Campus Lab1
  • Speaker: Ann-Sophie Mueller (Technical University Dresden, Germany)

Functional Benefit of Plastic Spiking Neural Networks for Information Processing

  • Date:  June 17, 2019
  • Venue: OIST Campus Lab1
  • Speaker: Dr. Xuhui Huang (Research Center for Brain-inspired Intelligence, Institute of Automation, Chinese Academy of Sciences)

Neural encoding and decoding of auditory cortex during perceptual decision making

  • Date:  August 15, 2019
  • Venue: OIST Campus Lab1
  • Speaker: Dr. Akihiro Funamizu (Zador Laboratory, Cold Spring Harbor Laboratory)

Active Sensing and other endogenously generated processes

  • Date:  September 12, 2019
  • Venue: OIST Campus Lab1
  • Speaker: Dr. Pedro Maldonado A (Departamento de Neurociencia & BNI, Facultad de Medicina Universidad de Chile) (School of Informatics, University of Edinburgh)

Spatio-Temporal Cortical Spike Patterns in Motor Cortex

  • Date:  September 12, 2019
  • Venue: OIST Campus Lab1
  • Speaker: Dr. Sonja Grün (Institute of Neuroscience and Medicine (INM-6 and INM-10) 
    Institute for Advanced Simulation (IAS-6), Jülich Research Center, Germany Theoretical Systems Neurobiology, RWTH Aachen University, Germany)

Learning from social data to study human behaviour

  • Date:  November 1, 2019
  • Venue: OIST Campus Lab1
  • Speaker: Dr. Scott A.Hale (Oxford Internet Institute, University of Oxford)

Multiple independent goal-directed learning for emergent adaptive behaviour in robots

  • Date:  November 20, 2019
  • Venue: OIST Campus Lab1
  • Speaker: Dr. Danish Shaikh (University of Southern Denmark Embodied Systems for Robotics and Learning, The Maersk Mc-Kinney Moller Institute)

Open cortical multi-area model as research platform

  • Date:  December 9, 2019
  • Venue: Venue: OIST Campus Center bldg.
  • Speaker:  Dr. Markus Diesmann (Institute of Neuroscience and Medicine (INM-6), Computational and Systems Neuroscience & Institute for Advanced Simulation (IAS-6), Theoretical Neuroscience & JARA-Institut Brain structure-function relationships (INM-10).)

Introduction to NEST and NEST3 features

  • Date:  December 10, 2019
  • Venue: OIST Campus Center bldg.
  • Speaker:Dr. Dennis Terhorst (Institute of Neuroscience and Medicine (INM-6), Computational and Systems Neuroscience & Institute for Advanced Simulation (IAS-6), Theoretical Neuroscience & JARA-Institut Brain structure-function relationships (INM-10))

6.2 Events

The 6th Research Area Meeting "Artificial Intelligence and Brain Science"

  • Date: May 14, - 15, 2019
  • Venue: Tamagawa University
  • Sponsor: Kakenhi Project on Artificial Intelligence and Brain Science

NC, IBISML, MPS, BIO Joint Technical Workshop

  • Date: June 17- 19, 2019
  • Venue: OIST Auditrium, Conference Center
  • Co-sponsors:
    Japanese Neural Network Society
    IEEE Computational Intelligence Society Japan Chapter
    Okinawa Institute of Science and Technology

The 7th Research Area Meeting "Artificial Intelligence and Brain Science"

  • Date: December 20, - 21, 2019
  • Venue: Hitotsubashi University
  • Sponsor: Kakenhi Project on Artificial Intelligence and Brain Science

7. Other

Prof. Kenji Doya received Outstanding Achievement Award from the Asia-Pacific Neural Network Society.

Prof. Kenji Doya received Academic Award of Japanese Neural Networks Society.

Paavo Parmas received Best Reviewer Award of NeurIPS Conference 2019 for his excellent contribution as a reviewer.