Theory ofReinforcementLearningFrom Pavlov to Rescorla and Wagner. Understanding the functional role ofdopamineneurons, however, requires more than a knowledge of brain circuitry; it also requires an understanding of the classes of computational algorithms in whichdopamineneurons seem to participate.
In the present study, we constructed a closed neural circuit model based on this conjecture and tried to simultaneously explain the observed motivational effects ofdopaminereceptor antagonists and the suggested roles ofdopamineinreinforcementlearningby simulating two behavioral tasks. Materials and Methods Simulated saccade task.
These results are in line with the idea thatdopamineregulates decision thresholds duringreinforcementlearning,andmay help to bridge action selection and response vigor accounts ofdopamine.
MIT researchers have found surprising patterns ofdopaminesignaling that suggest neuroscientists may need to refine their model of howreinforcementlearningoccurs in the brain.
Reinforcementlearningis the problem of predicting and maximizing long-term reward. Computer scientists recognized that this problem could be solved by updating predictions and action policies based on prediction errors (discrepancies between observed and expected reward). Remarkably, a similar strategy appears to be used by the basal ganglia, wheredopamineprovides the prediction errors for ...
Human-level control through deep ofdopamineprediction error: Implications of a parallel circuitreinforcementlearning. Nature, 518(7540), 529 -533. architecture in the basal ganglia.
INTRODUCTION This chapter provides an overview ofreinforcementlearningandtemporal differencelearningandrelates these topics to the firing properties of midbraindopa-mineneurons. First, we review the Rescorla Wagnerlearningrule and basiclearningphenomena, such as blocking, which the rule explains. Then we introduce the basic functional anatomy of thedopaminesystem and review ...
