Large Margin Feature Weighting Method via Linear Programming

Authors:
Bo Chen;Hongwei Liu;Jing Chai;Zheng Bao
Affiliations:
Xidian University , Xi'an;Xidian University, Xian;Xidian University, Xi'an;Xidian University, Xi'an
Venue:
IEEE Transactions on Knowledge and Data Engineering
Year:
2009

Citing 0
Cited 7

Generalized re-weighting local sampling mean discriminant analysis

Pattern Recognition
Modeling recognizing behavior of radar high resolution range profile using multi-agent system

WSEAS Transactions on Information Science and Applications
A novel hypothesis-margin based approach for feature selection with side pairwise constraints

Neurocomputing
Optimizing search engines results using linear programming

Expert Systems with Applications: An International Journal
Radar HRRP recognition based on discriminant information analysis

WSEAS Transactions on Information Science and Applications
On the evolutionary optimization of k-NN by label-dependent feature weighting

Pattern Recognition Letters
Robust feature selection based on regularized brownboost loss

Knowledge-Based Systems

Quantified Score

Hi-index	0.00

Visualization

Abstract

The problem of feature selection is a difficult combinatorial task in machine learning and of high practical relevance. In this paper, we consider feature selection method for multimodally distributed data, and present a large margin feature weighting method for k-nearest neighbor (kNN) classifiers. The method learns the feature weighting factors by minimizing a cost function, which aims at separating different classes by large local margins and pulling closer together points from the same class, based on using as few features as possible. The consequent optimization problem can be efficiently solved by Linear Programming. Finally, the proposed approach is assessed through a series of experiments with UCI and microarray data sets, as well as a more specific and challenging task, namely, radar high-resolution range profiles (HRRP) automatic target recognition (ATR). The experimental results demonstrate the effectiveness of the proposed algorithms.