Klaas Seinhorst is the guest speaker at this ACLC seminar, the title and abstract of this lecture are now available.
|Date||1 June 2018|
|Time||15:15 - 16:30|
An extensive body of research suggests that the complexity of a data set is negatively correlated with its learnability. An obvious strategy to improve learnability, then, is reducing complexity, for instance by regularization, i.e. elimination of exceptions. I will present results from two implicit learning experiments that investigate patterns of phonological features. In the first experiment, participants displayed regularizing behavior; in the other, the reduction of complexity turned out to have its limits.