A Learning Criterion for Stochastic Rules

Machine Learning

Volumn 9, Issue 2, 1992, Pages 165-203

  Yamanishi, Kenji ^a  

^a NEC CORPORATION   (Japan)

Author keywords

Learning from examples; MDL principle; PAC model; sample complexity; stochastic decision lists; stochastic decision trees; stochastic rules

Indexed keywords

EID: 0000819141     PISSN: 08856125     EISSN: 15730565     Source Type: Journal    
DOI: 10.1023/A:1022641132503     Document Type: Article

References (41)

1. On the computational complexity of approximating distributions by probabilistic automata
2. A new look at the statistical model identification
3. Minimum complexity density estimation
4. Occam's razor
5. Learnability and Vapnik-Chervonenkis dimension
6. Learning the distribution in the extended PAC model
7. A general lower bound on the number of examples needed for learning
8. Decision theoretic generalizations of the PAC learning model
9. Efficient distribution-free learning of probabilistic concepts
10. Some conditions for consistency and uniform consistency of statistical procedures
11. A lower bound for discrimination in terms of variation
12. Efficient unsupervised learning
13. Computational limitation on learning from examples
14. Inferring decision trees using the minimum description length criterion
15. Modeling by shortest data description
16. A universal prior for integers and estimation by minimum description length
17. Universal coding, information, prediction, and estimation
18. Stochastic complexity and modeling
19. Learning decision lists
20. The Bayes Laplace statistic of the multinomial distributions
21. Types of noise in data for concept learning
22. A formal theory of inductive inference
23. A theory of the learnable
24. Learning disjunctions of conjunctions
25. On the uniform convergence of relative frequencies of events to their probabilities
26. Inductive inference and learning criterion of stochastic classification rules with hierarchical parameter structures