23 Jun Categorization
Posted at 18:05h
in Uncategorized
We do it every day: correctly classifying people, objects, words. What seems easy for humans is computationally very complex. Together with Jonathan D. Nelson and Björn Meder (both Center for Adaptive Behavior and Cognition, MPIB Berlin) I hypothesized and found that a relatively simple statistical principle called class-conditional independence (aka naive Bayes) describes human classification learning.