Output details
11 - Computer Science and Informatics
University of York
Article title
Kernel Bandwidth Estimation for Nonparametric Modeling
Type
D - Journal article
Title of journal
Systems, Man, and Cybernetics, Part B: Cybernetics, IEEE Transactions on
Article number
-
Volume number
39
Issue number
6
First page of article
1543
ISSN of journal
1083-4419
Year of publication
2009
Number of additional authors
1
Additional information
<24>A novel non-parametric Bayesian approach is introduced for sampling the kernel bandwidth parameter used in various non-parametric statistical methods. The paper applies this method to a number of well known problems including space-scale analysis, mean-shift analysis and quantum clustering, where it improves efficiency. The method is also shown to improve the efficiency of various non-parametric image processing applications including motion estimation, non-rigid image registration, anomaly detection in hyperspectral images, and ladar image segmentation.
Interdisciplinary
-
Cross-referral requested
-
Research group
H - Computer Vision and Pattern Recognition
Citation count
11
Proposed double-weighted
No
Double-weighted statement
-
Reserve for a double-weighted output
No
Non-English
No
English abstract
-
