15 Relational Planning, Learning, and Probabilistic Reasoning

15.4 Review

The following are the main points you should have learned from this chapter:

  • Relational representations are used when an agent requires models to be given or learned before it which individuals it will encounter.

  • Many of the representations in earlier chapters can be made relational.

  • The situation calculus represents time in terms of the action of an agent, using the init constant and the do function.

  • Event calculus allows for continuous and discrete time and axiomatizes what follows from the occurrence of events.

  • Inductive logic programming can be used to learn relational models, even when the values of features are meaningless names.

  • Collaborative filtering can be used to make predictions about instances of relations from other instances by inventing hidden properties.

  • Plate models and the independent choice logic allow for the specification of probabilistic models before the individuals are known.