1 min read

Temporal Difference Model

There is a list of Temporal Difference models that can be used to model animals' learning processes:

  • TD(0)
  • Q-learning
  • Expected SARSA
  • TD(λ)

They all have in common: that predictions are getting adjusted on every step with fresh predictions that are closer to the outcome and more accurate. Before the final outcome is known.

Suppose you wish to predict the weather for Saturday, and you have some model that predicts Saturday's weather, given the weather of each day in the week. In the standard case, you would wait until Saturday and then adjust all your models. However, when it is, for example, Friday, you should have a pretty good idea of what the weather would be on Saturday – and thus be able to change, say, Saturday's model before Saturday arrives


Temporal difference learning - Wikipedia Animal cognition - Wikipedia