A dynamic selection ensemble method for target recognition based on clustering and randomized reference classifier
In order to improve the generalization ability and recognition efficiency of the maritime surveillance radar, a novel selection ensemble technique, termed KMRRC, based on k-medoids clustering and random reference classifier (RRC) is proposed. By disturbing the training set base classifiers are generated, which are then divided into several clusters based on pairwise diversity metrics, finally the RRC model is used to select several most competent classifiers from each cluster to classify each query object. The performance of KMRRC is compared against nine ensemble learning methods using a self-built high range resolution profile (HRRP) data set and twenty UCI databases. The experimental results clearly show the KMMRRC’s feasibility and effectiveness. In addition, the influence of the selection of diversity measures is studied concurrently.
KeywordsK-medoids clustering Randomized reference classifier Dynamic selection ensemble Target recognition
This work is supported by Grant No. (61401493) from the National Natural Science Foundation of China and No.(9140A01010415JB11002) from the National Ministries Foundation of China.
- 6.Azizi N, Farah N, Ennaji A (2012) New dynamic classifiers selection approach for handwritten recognition. In: Proceedings of international conference on artificial neural networks and machine learning, Springer, pp 189–196Google Scholar
- 11.Cruz R M O, Cavalcanti G D C, Ren TI (2011) A method for dynamic ensemble selection based on a filter and an adaptive distance to improve the quality of the regions of competence. In: Proceedings of International Joint Conference on Neural Networks, IEEE, pp 1126–1133Google Scholar
- 12.Azizi N, Farah N, Khadir T (2015) Multi stage dynamic ensemble selection using heterogeneous learning algorithms: application on classification problems. Int J Knowl Manag Stud 142(1):98–104Google Scholar
- 15.Krijthe JH, Ho TK, Loog M (2012) Improving cross-validation based classifier selection using meta-learning. In: Proceedings of international conference on pattern recognition, IEEE, pp 2873–2876Google Scholar
- 20.Zenobi G, Cunningham P (2001) Using Diversity in Preparing Ensembles of Classifiers Based on Different Feature Subsets to Minimize Generalization Error. In: Proceedings of European conference on machine learning, Springe, pp 576–587Google Scholar
- 23.Cameron WL, Leung LK (1990) Feature motivated polarization scattering matrix decomposition. In: Proceedings of international conference on radar, IEEE, pp 549–557Google Scholar
- 27.Berizzi F, Martorella M, Capria A et al.(2008) H/α polarimetric features for man-made target classification. In: Proceedings of international conference on radar conference, IEEE, pp 1–6Google Scholar
- 29.Guo L (2009) Wideband radar target polarimetric feature extraction and recognition method based on kernel method. Dissertation, National University of Defense Technology. (in chinese)Google Scholar
- 30.Fan XM, Hu SL, He JB (2016) Feature extraction and selection of Full Polarization HRRP in the target recognition process of maritime surveillance radar. J Electrs Inf Technol 38(12):3261–3268Google Scholar
- 33.Balabantaray RC, Sarma C, Jha M (2015) Document clustering using K-means and K-medoids. Int J Knowl Based Comput Syst 1:7–13Google Scholar
- 34.Azizi N, Farah N (2012) From static to dynamic ensemble of classifiers selection: application to Arabic handwritten recognition. IOS press, USA, pp 279–288Google Scholar