Computational limitations on learning from examples
Leonard Pitt, Leslie Gabriel Valiant · Journal of the ACM · 1988
The computational complexity of learning Boolean concepts from examples is investigated. It is shown for various classes of concept representations that these cannot be learned feasibly in a distribution-free sense unless R = NP. These classes include (a) disjunctions of two monomials, (b) Boolean threshold functions, and (c) Boolean formulas in which each variable occurs at most once. Relationships between learning of heuristics and finding approximate solutions to NP-hard optimization problems are given.
Cited by 515 works. doi:10.1145/48014.63140 Open-access PDF
Authors
References
- Computers and Intractability: A Guide to the Theory of NP-Completeness
- A Theory of the Learnable
- A theory of the learnable
- How to construct random functions
- Inductive Inference: Theory and Methods
- Occam's Razor
- Computational Complexity of Probabilistic Turing Machines
- A Greedy Heuristic for the Set-Covering Problem
- Proceedings 35th Annual Symposium on Foundations of Computer Science
- Classifying learnable geometric concepts with the Vapnik-Chervonenkis dimension