Computational Psychiatry

Through temporal difference (TD) learning, an agent learns from future rewards and backpropagates prediction errors by updating value estimates (continually updating beliefs about future rewards as it approaches the future). It is a core concept of model-free reinforcement learning.

The Rescorla-Wagner model captures key aspects of classical conditioning(Pavlovian experiment). It is based on a simple linear equation that predicts the reward associated with a stimulus.

计算精神病学领域较常见的三类研究范式总结。

Install Hierachical Drift Diffusion Model via Conda.

Test