Publications

BibTex format

@article{Gillies:2015:10.1016/j.patcog.2015.11.015,
author = {Gillies, DF and Liu, R},
doi = {10.1016/j.patcog.2015.11.015},
journal = {Pattern Recognition},
pages = {73--86},
title = {Overfiting in linear feature extraction for classificationof high-dimensional image data},
url = {http://dx.doi.org/10.1016/j.patcog.2015.11.015},
volume = {53},
year = {2015}
}

Download

RIS format (EndNote, RefMan)

TY  - JOUR
AB  - Overfitting has been widely studied in the context of classification and regression. In this paper, we study the overfitting in the context of dimensionality reduction. We show that the conventional wisdom of improving classification performance by maximising inter-class discrimination is not valid for high-dimensional datasets, and can lead to severe overfitting. In particular, we prove the theoretical existence of perfectly discriminative subspace projections, and show that for datasets with very high input dimensionality, inter-class discrimination should be reduced rather than maximised. This naturally leads to a simple dimensionality reduction technique, which we call Soft Discriminant Maps, which we use to show a direct relationship between the classification performance and the level of inter-class discrimination of feature extractors. Moreover, Soft Discriminant Maps consistently exhibit better classification performance than other comparable techniques.
AU  - Gillies,DF
AU  - Liu,R
DO  - 10.1016/j.patcog.2015.11.015
EP  - 86
PY  - 2015///
SN  - 1873-5142
SP  - 73
TI  - Overfiting in linear feature extraction for classificationof high-dimensional image data
T2  - Pattern Recognition
UR  - http://dx.doi.org/10.1016/j.patcog.2015.11.015
UR  - http://hdl.handle.net/10044/1/33407
VL  - 53
ER  -

Download

ProfessorDuncanGillies

Contact

Location

Summary

Citation

BibTex format

RIS format (EndNote, RefMan)