CN302ADVANCEDCOMPUTATIONAL NEUROSCIENCE

Learning, Decision Making and Motor Control

How brains learn from reward, choose between options and control movement.

Course Information

Course Information
ModeInstructor-led
TermFA1 · Aug–Sep · proposed
Duration8 weeks · about 40 learner hours
Credits1 credit
PrerequisitesCN201.
Who it is forLearners with the Computational Neuroscience certificate or equivalent.
FacultyNeurovidya faculty
AssessmentAssignments and a graded project.

About the course

Reinforcement learning, decision models and motor control, connected to the basal ganglia, cerebellum and cortex, and to the disorders that follow when these circuits fail.

Learning outcomes

  • Model learning with temporal-difference RL and relate it to dopamine
  • Build basal ganglia models of action selection
  • Fit drift–diffusion models to choices and reaction times
  • Explain optimal feedback control and cerebellar learning
  • Relate model failures to Parkinson’s disease and other disorders

Course modules

7 modules

  • TD learning and dopamine

Labs and projects: Simulation projects in Python.

Books and readings: Chakravarthy & Moustafa, Computational Neuroscience Models of the Basal Ganglia (Springer); Sutton & Barto, Reinforcement Learning