An improved branch & bound algorithm in feature selection

Authors:
Zhenxiao Wang;Jie Yang;Guozheng Li
Affiliations:
Institute of Pattern Recognition & Image processing, Shanghai Jiaotong University, Shanghai, China;Institute of Pattern Recognition & Image processing, Shanghai Jiaotong University, Shanghai, China;Institute of Pattern Recognition & Image processing, Shanghai Jiaotong University, Shanghai, China
Venue:
RSFDGrC'03 Proceedings of the 9th international conference on Rough sets, fuzzy sets, data mining, and granular computing
Year:
2003

Citing 3
Cited 1

Introduction to statistical pattern recognition (2nd ed.)

Introduction to statistical pattern recognition (2nd ed.)
Feature selection toolbox software package

Pattern Recognition Letters - In memory of Professor E.S. Gelsema
A Branch and Bound Algorithm for Feature Subset Selection

IEEE Transactions on Computers

U-curve: A branch-and-bound optimization algorithm for U-shaped cost functions on Boolean lattices applied to the feature selection problem

Pattern Recognition

Quantified Score

Hi-index	0.00

Visualization

Abstract

The Branch & Bound (B&B) algorithm is a globally optimal feature selection method. The high computational complexity of this algorithm is a well-known problem. The B&B algorithm constructs a search tree, and then searches for the optimal feature subset in the tree. Previous work on the B&B algorithm was focused on how to simplify the search tree in order to reduce the search complexity. Several improvements have already existed. A detailed analysis of basic B&B algorithm and existing improvements is given under a common framework in which all the algorithms are compared. Based on this analysis, an improved B&B algorithm, BBPP+, is proposed. Experimental comparison shows that BBPP+ performs best.