What are the computations of the cerebellum, the basal ganglia and the cerebral cortex?

By Kenji Doya. 1999

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Citations in the corpus

A Comprehensive Survey of Continual Learning: Theory, Method and Application (2023, background): “In contrast, the neocortex-dependent memories are typically long-term and generalizable, where the representations can be progressively acquired in an unsupervised fashion [102], [103] and exhibit orthogonality [119], [463], consistent with the advantages of self-supervised and pretrained representa”
Toward an Integration of Deep Learning and Neuroscience (2016, background): “Areas like the basal ganglia are involved in reinforcement learning and gating of discrete decisions (Doya, 1999; Sejnowski and Poizner, 2014).”
TD models of reward predictive responses in dopamine neurons (2002, background): “Recently, it has been speculated that certain cerebellar functions that can be mimicked with internal model approaches influence dopamine neuron activity (Doya, 1999).”
How Working Memory and the Cerebellum Collaborate to Produce Creativity and Innovation (2007, unknown): “…and cognitive processes, including working memory (e.g., Akshoomoff et al., 1997; Cabeza & Nyberg, 2000; Chein et al., 2003; Desmond & Fiez, 1998; Doya, 1999; Houk & Wise, 1995; Haruno et al., 1999; Haruno, Wolpert, & Kawato, 2001; Imamizu et al., 2003; Ito, 1993, 1997; Ivry, 1997; Leiner &…”
Supervised Learning in Spiking Neural Networks with ReSuMe: Sequence Learning, Classification, and Spike Shifting (2010, background): “This surprising result points out that spiking neurons can potentially be applied to forecast the behavior (ﬁring times) of other reference neurons or networks.”
Introduction to spiking neural networks: Information processing, learning and applications. (2011, background): “…is believed to be utilized by the neural motor centers to form internal representations of the body and the environment (Kawato and Gomi 1992a,b, Shidara et al. 1993, Miall and Wolpert 1996, Wolpert et al. 1998) or for behavioral simulations and the encapsulation of learned skills (Doya 1999).”
Metalearning and neuromodulation (2002, background): “Furthermore, the cerebellum seems to provide internal models of the environmental dynamics (Kawato, 1999), which can be helpful in longterm prediction of the future events and rewards (Doya, 1999).”
The Cerebellum, Sensitive Periods, and Autism (2014, background): “Due to these markedly contrasting architectures, it has been proposed that the cerebellar and cerebral cortex differ in their learning algorithm, with the cerebellum performing supervised learning, while the cerebral cortex performs unsupervised learning (Doya, 1999).”
Cerebellar supervised learning revisited: biophysical modeling and degrees-of-freedom control. (2011, background): “About ten years ago, Kenji Doya proposed that the cerebellum, cerebral cortex, and the basal ganglia implement supervised, unsupervised, and reinforcement learning algorithms, respectively, mainly based on system-level data and previous computational models [8].”
Action understanding as inverse planning. (2009, background): “…the forward problems of planning and reinforcement learning are often divided into model-based and model-free approaches (Daw, Niv, & Day-an, 2005; Doya, 1999; Sutton & Barto, 1998), and there is evidence that the brain has systems corresponding to both (Dickinson, 1985; Schultz et al., 1997).”
Policy Gradient Coagent Networks (2011, background): “Methods for solving sequential decision problems with delayed reward, where the problems are formulated as Markov decision processes (MDPs), have been compared to the learning mechanisms of animal brains [3] [4] [9] [10] [13] [20] [22].”
Goal-directed decision making as probabilistic inference: a computational framework and potential neural correlates. (2012, background): “Over time, this early view of habit and goal directedness as mutually exclusive accounts of behavior has given way to a more inclusive multiple-systems account, under which habitual and goal-directed control coexist as complementary mechanisms for action selection (Daw, Niv, & Dayan, 2005; Dayan, 20”
Intrinsically motivated action-outcome learning and goal-based action recall: a system-level bio-constrained computational model. (2013, background): “The BG are also a key brain structure important for reinforcement (trial-and-error) learning processes (Houk et al., 1995; Doya, 1999; Brown et al., 1999; Joel et al., 2002).”