Human-level concept learning through probabilistic program induction
Abstract
Not only do children learn effortlessly, they do so quickly and with a remarkable ability to use what they have learned as the raw material for creating new stuff. Lake et al. describe a computational model that learns in a similar fashion and does so better than current deep learning algorithms. The model classifies, parses, and recreates handwritten characters, and can generate new letters of the alphabet that look “right” as judged by Turing-like tests of the model's output in comparison to what real humans produce.
Document Details
- Document Type
- Pub Defense Publication
- Publication Date
- Dec 11, 2015
- Source ID
- 10.1126/science.aab3050
Entities
People
- Brenden M. Lake
- Joshua B. Tenenbaum
- Ruslan Salakhutdinov
Organizations
- Army Research Office
- Canadian Institute for Advanced Research
- Massachusetts Institute of Technology
- Natural Sciences and Engineering Research Council
- New York University
- Office of Naval Research
- University of Toronto