Projection-based approximation and a duality with kernel methods
angular smoothnessapproximationconvergencecurse of dimensionalitydimensionality reductionGaussian measureharmonic functionsHermite polynomialskernel regressionL2-approximation errorLaplacianlocal averagingminimax rate ofnoisy dataProjection pursuit regressionradial functionsrates ofsmoothnesstwo-dimensional problem
Projection pursuit regression (PPR) and kernel regression (KR) are methods for estimating a smooth function of several variables from noisy data. The PPR method was introduced by \textit{J. H. Friedman} and \textit{W. Stuetzle} [J. Am. Stat. Assoc. 76, 817-823 (1981)] and the KR-method was discussed, for instance, by \textit{C. J. Stone} [Ann. Stat. 5, 595-645 (1977; Zbl 0366.62051)] and \textit{G. Collomb} [Int. Stat. Rev. 49, 75-93 (1981; Zbl 0471.62039)]. The KR-method is essentially a method based on local averaging and it is known that the performance of this method is poor in high dimensions. Earlier examples have suggested that the PPR approach performs better than the KR approach. Here the authors investigate the following questions: For what sorts of regression functions is this true? When and by how much do PPR methods reduce the curse of dimensionality? They consider a two-dimensional problem and show that both the methods are complementary in nature. For a given function, if one method offers a dimensionality reduction, the other does not.
- Identifying linear combinations of ridge functions
- Universal approximation by hierarchical fuzzy systems
- On using prototype reduction schemes to optimize kernel-based nonlinear subspace methods.
- Ridgelets: estimating with ridge functions
- A note on universal approximation by hierarchical fuzzy systems
- Some problems in the theory of ridge functions
- A note on continuous sums of ridge functions
- Fuzzy projection pursuit density estimation by eigenvalue method
- Application of projection pursuit learning to boundary detection and deblurring in images
- Nonlinear black-box models in system identification: Mathematical foundations
- Linearized two-layers neural networks in high dimension
- Data-independent random projections from the feature-map of the homogeneous polynomial kernel of degree two
- Dual bases and discrete reproducing kernels: a unified framework for RBF and MLS approximation
- Characterization of an extremal sum of ridge functions
- Linear and nonlinear methods of relief approximation
- On the error of approximation by ridge functions with two fixed directions
- scientific article; zbMATH DE number 4028654 (Why is no real title available?)
- L2Version Of The Double Kernel Method
- USING THE MUTUAL INFORMATION COEFFICIENT TO IDENTIFY LAGS IN NONLINEAR MODELS
- On function recovery by neural networks based on orthogonal expansions
- Distribution-specific hardness of learning neural networks
- Comparing Methods for Multivariate Nonparametric Regression
- Approximation by sums of ridge functions with fixed directions
- Probabilistic partition of unity networks for high‐dimensional regression problems
- Approximation of curve-based sleeve functions in high dimensions
- On the representability of a continuous multivariate function by sums of ridge functions
- On the representability of a smooth function by sums of generalized ridge functions
- Learning non-parametric basis independent models from point queries via low-rank methods
- Relative efficiencies of model robust estimators in two-dimensional situations
- A representation problem for smooth sums of ridge functions
- On Kolmogorov's representation of functions of several variables by functions of one variable
- Gibbs sampling, exponential families and orthogonal polynomials
- On a smoothness problem in ridge function representation
- Nonparametric estimation when data on derivatives are available
This page was built for publication: Projection-based approximation and a duality with kernel methods
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q913418)
