• Publications
  • Influence
A Distribution-Free Theory of Nonparametric Regression
Why is Nonparametric Regression Important? * How to Construct Nonparametric Regression Estimates * Lower Bounds * Partitioning Estimates * Kernel Estimates * k-NN Estimates * Splitting the Sample *Expand
  • 1,459
  • 215
  • PDF
Methods of combining multiple classifiers and their applications to handwriting recognition
TLDR
Possible solutions to the problem of combining classifiers can be divided into three categories according to the levels of information available from the various classifiers. Expand
  • 2,337
  • 130
  • PDF
Rival penalized competitive learning for clustering analysis, RBF net, and curve detection
TLDR
An algorithm called rival penalized competitive learning (RPCL) is proposed. Expand
  • 634
  • 61
Learning and Design of Principal Curves
TLDR
We define principal curves as continuous curves of a given length which minimize the expected squared distance between the curve and points of the space randomly chosen according to a given distribution. Expand
  • 337
  • 52
  • PDF
Fast SVM training algorithm with decomposition on very large data sets
TLDR
We introduce a parallel optimization step to quickly remove most of the nonsupport vectors, where block diagonal matrices are used to approximate the original kernel matrix so that the problem can be split into hundreds of subproblems which can be solved more efficiently. Expand
  • 235
  • 15
On the Strong Universal Consistency of Nearest Neighbor Regression Function Estimates
Two results are presented concerning the consistency of the k-nearest neighbor regression estimate. We show that all modes of convergence in L 1 (in probability, almost sure, complete) are equivalentExpand
  • 223
  • 15
  • PDF
Piecewise Linear Skeletonization Using Principal Curves
TLDR
Proposes an algorithm to find piecewise linear skeletons of handwritten characters by using principal curves. Expand
  • 221
  • 14
  • PDF
Image denoising using neighbouring wavelet coefficients
TLDR
We propose one wavelet image thresholding scheme by incorporating neighbouring coefficients, namely NeighShrink, which is better than the Wiener filter and the conventional wavelet denoising approaches: Visu Shrink and SUREShrink. Expand
  • 102
  • 14
  • PDF
Principal curves: learning, design, and applications
The subjects of this thesis are unsupervised learning in general, and principal curves in particular. Principal curves were originally defined by Hastie [Has84] and Hastie and Stuetzle [HS89]Expand
  • 64
  • 14
Robust Estimation for Range Image Segmentation and Reconstruction
TLDR
This correspondence presents a segmentation and fitting method using a new robust estimation technique using a compressed histogram method which can tolerate more than 80% of outliers. Expand
  • 120
  • 12
...
1
2
3
4
5
...