Overview:This lecture series is part of PSYC3211, and covers a range of topics in computational modelling of higher order cognition.
Associative learning via the Rescorla-Wagner rule. Connection to other error driven learning rules. Using networks as classifiers. More complex networks and pattern matching.
Introduction to Bayesian reasoning. A model for judging coincidences. Comments on conservative belief upating. A model of the perceptual magnet effect. Bayesian program induction for concept learning.
Semantic priming and spreading activation. The small world of words project. Local network structure. Predicting remote associations. Structure of semantic networks. Developmental trajectory
Galton's vox populi. Surowiecki's criteria. Wisdom of crowds for ranking data. Example from category learning. Wisdom of crowds in combinatorial optimisation problems. Compensating for strategic behaviour. Application in forensic science
The iterated learning paradigm. Theoretical argument that it reveals inductive biases. Illustration with function learning task. Limitations when individual differences exist. Cumulative cultural evolution in a language game.