Radar HRRP target recognition based on Gradient Boosting Decision Tree
Authors:
- Suixue WangSWSuixue Wang
- Jinbing LiJLJinbing Li
- Yanhua WangYWYanhua Wang
- Yang LiYLYang Li
- 1 October 2016
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1013-1017
Abstract
Radar High Resolution Range Profile (HRRP) represents the radial distribution of target scattering centers on radar line of sight, which infers something about target structure signatures. In this paper, the method of HRRP target recognition based on Gradient Boosting Decision Tree (GBDT) is proposed and the method how to choose model parameters is investigated. The experimental results from measurement data have shown that GBDT can achieve better recognition results and high calculation efficiency than SVM and Naive Bayes classifier.Keywords
This publication has 7 references indexed in Scilit:
- The Radar Target Recognition Research Based on Improved Neural Network Algorithm2014
- Radar HRRP statistical recognition based on hypersphere modelSignal Processing, 2008
- A working guide to boosted regression treesJournal of Animal Ecology, 2008
- Features for high resolution radar range profile based ship classification2003
- Introduction to radar systems third edition [Book Review]IEEE Aerospace and Electronic Systems Magazine, 2001
- Optimal polarimetric processing for enhanced target detectionIEEE Transactions on Aerospace and Electronic Systems, 1993
- Induction of decision treesMachine Learning, 1986