Computational Neuropsychiatry

Question 1

How does the Rescorla-Wagner model propose we update our expectations?

Answer 1

Prediction errors:
Difference between what you expect and observe

This is weighted - learning rate

Question 2

What is the equation for prediction in the Rescorla-Wagner model?

Answer 2

Expectation - (rate x error)

Question 3

What has been shown about schizophrenia and learning rate?

Answer 3

Learn more slowly than healthy controls

(Could be overemphasis on top down control and not sensory feedback)

Question 4

What has been shown about anxiety and learning rate?

Answer 4

People with high anxiety have higher learning rates from negative outcomes

Question 5

What are the two decision making systems?

Answer 5

Goal directed
Habitual

Question 6

What symptom across neuropsychiatric diagnosed is linked to habitual decision-making?

Answer 6

Compulsivity and intrusive thought

Question 7

How can computational methods be applied to neuropsychiatry?

Answer 7

Theory-driven - models of mechanisms

Data-driven - distinguishing people by diagnosis based on data