Abstract
Feature extraction has been an important research topic in pattern classification and has been studied extensively by many researchers. Most of the conventional feature extraction methods are performed using a criterion function defined between two classes or a global function. Although these methods work relatively well in most cases, it is generally not optimal in any sense for multiclass problems. In order to address this problem, we propose a method to optimize feature extraction for multiclass problems. We first investigated the distribution of classification accuracies of multiclass problems in the feature space and found that there exist much better feature sets that the conventional feature extraction algorithms fail to find. Then we proposed an algorithm that finds such features. Experiments with remotely sensed data show that the proposed algorithm consistently provides better performances compared with the conventional feature extraction algorithms.
| Original language | English |
|---|---|
| Pages (from-to) | 521-528 |
| Number of pages | 8 |
| Journal | IEEE Transactions on Geoscience and Remote Sensing |
| Volume | 39 |
| Issue number | 3 |
| DOIs | |
| Publication status | Published - 2001 Mar 1 |
Fingerprint
All Science Journal Classification (ASJC) codes
- Electrical and Electronic Engineering
- Earth and Planetary Sciences(all)
Cite this
}
Optimizing feature extraction for multiclass problems. / Choi, E.; Lee, C.
In: IEEE Transactions on Geoscience and Remote Sensing, Vol. 39, No. 3, 01.03.2001, p. 521-528.Research output: Contribution to journal › Article
TY - JOUR
T1 - Optimizing feature extraction for multiclass problems
AU - Choi, E.
AU - Lee, C.
PY - 2001/3/1
Y1 - 2001/3/1
N2 - Feature extraction has been an important research topic in pattern classification and has been studied extensively by many researchers. Most of the conventional feature extraction methods are performed using a criterion function defined between two classes or a global function. Although these methods work relatively well in most cases, it is generally not optimal in any sense for multiclass problems. In order to address this problem, we propose a method to optimize feature extraction for multiclass problems. We first investigated the distribution of classification accuracies of multiclass problems in the feature space and found that there exist much better feature sets that the conventional feature extraction algorithms fail to find. Then we proposed an algorithm that finds such features. Experiments with remotely sensed data show that the proposed algorithm consistently provides better performances compared with the conventional feature extraction algorithms.
AB - Feature extraction has been an important research topic in pattern classification and has been studied extensively by many researchers. Most of the conventional feature extraction methods are performed using a criterion function defined between two classes or a global function. Although these methods work relatively well in most cases, it is generally not optimal in any sense for multiclass problems. In order to address this problem, we propose a method to optimize feature extraction for multiclass problems. We first investigated the distribution of classification accuracies of multiclass problems in the feature space and found that there exist much better feature sets that the conventional feature extraction algorithms fail to find. Then we proposed an algorithm that finds such features. Experiments with remotely sensed data show that the proposed algorithm consistently provides better performances compared with the conventional feature extraction algorithms.
UR - http://www.scopus.com/inward/record.url?scp=0035275083&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=0035275083&partnerID=8YFLogxK
U2 - 10.1109/36.911110
DO - 10.1109/36.911110
M3 - Article
AN - SCOPUS:0035275083
VL - 39
SP - 521
EP - 528
JO - IEEE Transactions on Geoscience and Remote Sensing
JF - IEEE Transactions on Geoscience and Remote Sensing
SN - 0196-2892
IS - 3
ER -