Learning with prior information

IEEE Transactions on Automatic Control

Volumn 46, Issue 11, 2001, Pages 1682-1695

  Campi, M C ^b   Vidyasagar, M ^a,c  

^a IEEE   (India)

^b UNIVERSITY OF BRESCIA   (Italy)

^c TATA CONSULTANCY SERVICES   (India)

Author keywords

Bayesian learning; Dispersability; Learning theory; System identification

Indexed keywords

BAYESIAN LEARNING; DISPERSABILITY;

APPROXIMATION THEORY; CONTROL SYSTEM SYNTHESIS; IDENTIFICATION (CONTROL SYSTEMS); LEARNING ALGORITHMS; NONLINEAR CONTROL SYSTEMS; PROBABILITY;

LEARNING SYSTEMS;

EID: 0035508945     PISSN: 00189286     EISSN: None     Source Type: Journal    
DOI: 10.1109/9.964680     Document Type: Article

References (30)

1. Universal approximation bounds for superpositions of a sigmoidal function
2. Fat-shattering and the learnability of real-valued functions
3. Learnability by fixed distribution
4. Nonuniform learnability
5. Probability and measure
6. Learnability and the Vapnik-Chervonenkis dimension
7. Hinging hyperplanes for regression, classification and function approximation
8. Learning by canonical smooth estimation-Part I: Simultaneous estimation
9. Learning by canonical smooth estimation-Part II: Learning and choice of model complexity
10. A general lower bound on the number of examples needed for learning
11. Abstract inference
12. Decision theoretic generalizations of the PAC model for neural net and other learning applications
13. Bounds on the sample complexity of Bayesian learning using information theory and the VC dimension
14. Estimating average-case learning curves using Bayesian, statistical physics and VC dimension methods
15. ε-entropy and ε-capacity of sets in functional spaces
16. PAC-Bayesian theorems
17. Empirical processes: Theory and applications
18. A PAC analysis of a Bayesian estimator
19. A theory of the learnable
20. Uniform convergence of the frequencies of occurrence of events to their probabilities
21. On the uniform convergence of relative frequencies to their probabilities
22. Necessary and sufficient conditions for the uniform convergence of means of their expectations