Access
To read this story in full you will need to login or make a payment (see right).
News and Views
Nature 428, 378 (25 March 2004) | doi:10.1038/428378a
Open Innovation Challenges
-
Optimizing Sub-cellular Localization Tags
The Seeker is looking for methods to optimize sub-cellular localization tags for protein expression....
-
Methods of Modeling Adaptation in Populations
The analysis of adaptation with a population is a frequently encountered computational modeling scen...
- More Challenges
- Powered by:
nature jobs
Academic Surgical Pathologist
- Beth Israel Deaconess Medical Center (BIDMC) / Harvard Medical School
- Boston, MA
Senior Faculty Positions
- Torrey Pines Institute for Molecular Studies
- Port St. Lucie, FL
Learning theory: Past performance and future results
Carlo Tomasi1
Abstract
Learning from experience is hard, and predicting how well what we have learned will serve us in the future is even harder. The most useful lessons turn out to be those that are insensitive to small changes in our experience.
A hallmark of intelligent learning is that we can apply what we have learned to new situations. In the mathematical theory of learning, this ability is called generalization.
- Department of Computer Science, Duke University, Durham, North Carolina 27708, USA.
Email: tomasi@cs.duke.edu
To read this story in full you will need to login or make a payment (see right).
MORE ARTICLES LIKE THIS
These links to content published by NPG are automatically generated.
RESEARCH
General conditions for predictivity in learning theoryNature Letters to Editor (25 Mar 2004)
