Model selection and error estimation

Machine Learning

Volumn 48, Issue 1-3, 2002, Pages 85-113

  Bartlett, Peter L ^a   Boucheron, Stéphane ^b   Lugosi, Gábor ^c  

^a BIOwulf Technologies   (United States)

^b UFR 919 Laboratoire d'Informatique Pour la Mécanique et les Sciences de l'Ingénieur   (France)

^c UNIVERSITAT POMPEU FABRA   (Spain)

Author keywords

Concentration inequalities; Empirical penalties; Model selection; Penalization

Indexed keywords

CALCULATIONS; COMPUTATIONAL COMPLEXITY; ERROR ANALYSIS; ESTIMATION; INTEGRATION; MATHEMATICAL MODELS; MONTE CARLO METHODS; PATTERN RECOGNITION;

CONCENTRATION INEQUALITIES; EMPIRICAL PENALTIES; ERROR ESTIMATION; MODEL SELECTION;

LEARNING SYSTEMS;

EID: 0036643049     PISSN: 08856125     EISSN: None     Source Type: Journal    
DOI: 10.1023/A:1013999503812     Document Type: Article

References (51)

1. A new look at the statistical model identification
2. Logically smooth density estimation
3. Complexity regularization with application to artificial neural networks
4. Risk bounds for model selection via penalization
5. Minimum complexity density estimation
6. The sample complexity of pattern classification with neural networks: The size of the weights is more important than the size of the network
7. Generalization performance of support vector machines and other pattern classifiers
8. From model selection to adaptive estimation
9. Minimum contrast estimators on sieves: Exponential bounds and rates of convergence
10. A sharp concentration inequality with applications in random combinatorics and learning
11. Learning by canonical smooth estimation, Part I: Simultaneous estimation
12. Learning by canonical smooth estimation, Part II: Learning and choice of model complexity
13. Self bounding learning algorithms
14. Nonparametric maximum likelihood estimation by the method of sieves
15. Some limit theorems for empirical processes
16. Probability inequalities for sums of bounded random variables
17. An experimental and theoretical comparison of model selection methods
18. Rademacher penalties and structural risk minimization
19. Rademacher processes and bounding the risk of function learning
20. Radial basis function networks and complexity regularization in function learning
21. Model selection using rademacher penalization
22. Adaptive model selection using empirical complexities
23. Nonparametric estimation via empirical risk minimization
24. Concept learning using complexity regularization
25. p
26. Improved generalization through explicit optimization of margins
27. Some applications of concentration inequalities to statistics
28. On the method of bounded differences
29. Performance bounds for nonlinear time series prediction
30. Minimum complexity regression estimation with weakly dependent observations
31. A universal prior for integers and estimation by minimum description length
32. Boosting the margin: A new explanation for the effectiveness of voting methods
33. Estimating the dimension of a model
34. Structural risk minimization over data-dependent hierarchies
35. Convergence rate of sieve estimates
36. On the best constants in the Khintchine inequality
37. Concentration of measure and isoperimetric inequalities in product spaces
38. On the uniform convergence of relative frequencies of events to their probabilities
39. Measuring the VC-dimension of a learning machine
40. Sample based generalization bounds
41. An asymptotic property of model selection criteria
42. Information-theoretic determination of minimax rates of convergence