A Simple CW-SSIM Kernel-based Nearest Neighbor Method for Handwritten Digit Classification
Machine Learning
2010-09-06 v3
Abstract
We propose a simple kernel based nearest neighbor approach for handwritten digit classification. The "distance" here is actually a kernel defining the similarity between two images. We carefully study the effects of different number of neighbors and weight schemes and report the results. With only a few nearest neighbors (or most similar images) to vote, the test set error rate on MNIST database could reach about 1.5%-2.0%, which is very close to many advanced models.
Keywords
Cite
@article{arxiv.1008.3951,
title = {A Simple CW-SSIM Kernel-based Nearest Neighbor Method for Handwritten Digit Classification},
author = {Jiheng Wang and Guangzhe Fan and Zhou Wang},
journal= {arXiv preprint arXiv:1008.3951},
year = {2010}
}
Comments
16 pages, 11 figures, 1 table and 3 equations