Where and when: Thursday, Oct 28 at 2-3pm in 303S-561 or Google Meet (see the calendar invite for the link).
In this paper, we pose a challenge for AI researchers: to develop systems that learn in a human-like manner. We briefly review the history of machine learning, noting that early work made close contact with results from cognitive psychology but that this is no longer the case. We identify seven characteristics of human behavior that, if reproduced, would offer more effective ways to acquire expertise than statistical induction over massive training sets. We illustrate our points with examples from two domains – mathematics and driving – where people are effective learners. In closing, we suggest ways to encourage more research on human-like learning.
Get the paper here.
This paper has been proposed by Pat Langley.