On radial basis function nets and kernel regression: Statistical consistency, convergence rates, and receptive field size

Neural Networks

Volumn 7, Issue 4, 1994, Pages 609-628

  Xu, Lei ^a,b   Krzyzak, Adam ^c   Yuille, Alan ^a  

^a HARVARD UNIVERSITY   (United States)

^b PEKING UNIVERSITY   (China)

^c Department of Applied Human Sciences   (United States)

Author keywords

Convergence rate; Kernel regression estimator; Parzen window estimator; Radial basis function networks; Receptive field size; Statistical consistency; Universal approximation

Indexed keywords

CONVERGENCE OF NUMERICAL METHODS; ERRORS; ESTIMATION; LEAST SQUARES APPROXIMATIONS; NUMERICAL ANALYSIS; PROBABILITY; REGRESSION ANALYSIS; STATISTICAL METHODS;

APPROXIMATION ERROR; CONVERGENCE RATE; KERNEL REGRESSION ESTIMATOR; PARZEN WINDOW ESTIMATOR; RADIAL BASIS FUNCTION NETWORK; RECEPTIVE FIELD SIZE; STATISTICAL CONSISTENCY; UNIVERSAL APPROXIMATION; UPPER BOUND;

NEURAL NETWORKS;

EID: 0028341934     PISSN: 08936080     EISSN: None     Source Type: Journal    
DOI: 10.1016/0893-6080(94)90040-X     Document Type: Article

References (45)

1. Approximation and estimation bounds for artificial neural networks
2. Probability inequalities for the sums of independent random variable
3. Generalization properties of radial basis function
4. Multivariable functional interpolation and adaptive networks
5. Orthogonal least squares learning algorithm for radial basis function networks
6. On the almost everywhere convergence of non-parametric regression function estimates
7. 1 convergence of the kernel regression estimate
8. Neural networks and the bias/variance dilemma
9. Convergence rates of approximation by translates
10. Networks and the best approximation property
11. Networks and the best approximation property
12. Distribution-free consistency of kernel regression estimate
13. Layered neural networks with Gaussian hidden units s universal approximations
14. Probability inequalities for sums of bounded random variables
15. Approximation capabilities of multilayer feedforward networks
16. Multilayer feedforward networks are universal approximators
17. Information theoretic derivation of network architecture and learning algorithms
18. Sequential adaptation of radial basis function neural networks and its application to time-series prediction
19. Generalization capabilities of minimal kernel-based networks
20. The rates of convergence of kernel regression estimates and classification rules
21. On exponential bounds on the Bayes risk of the kernal classification rule
22. The pointwise rate of convergence of the kernel regression estimate
23. How receptive field parameters affect neural learning
24. Fast learning in networks of locallytuned processing units
25. Max likelihood competition in RBF networks
26. Universal approximation using radial-basis-function networks
27. Universal approximation using radial-basis-function networks
28. Learning by combining memorization and gradient descent
29. A theory of networks for approximation and learning
30. Networks for approximation and learning
31. Radial basis functions for multivariable interpolation: A review
32. Phoneme classification experiments using radial basis functions
33. Probabilistic neural networks
34. A fast robust learning algorithm for feed-forward neural networks
35. Connectionist nonparametric regression: Multilayer feedforward networks that can learn arbitrary mappings
36. Recent advances on techniques static feedforward networks with supervised learning
37. Rival penalized competitive learning for clustering analysis, RBF net and curve detection
38. On radial basis function nets and kernel regression: Approximation ability, convergence rate and receptive field size
39. A mathematical analysis of the motion coherence theory