A genetic algorithm based wrapper feature selection method for classification of hyperspectral images using support vector machine

The high-dimensional feature vectors of hyper spectral data often impose a high computational cost as well as the risk of "over fitting" when classification is performed. Therefore it is necessary to reduce the dimensionality through ways like feature selection. Currently, there are two kinds of feature selection methods: filter methods and wrapper methods. The former kind requires no feedback from classifiers and estimates the classification performance indirectly. The latter kind evaluates the "goodness" of selected feature subset directly based on the classification accuracy. Many experimental results have proved that the wrapper methods can yield better performance, although they have the disadvantage of high computational cost. In this paper, we present a Genetic Algorithm (GA) based wrapper method for classification of hyper spectral data using Support Vector Machine (SVM), a state-of-art classifier that has found success in a variety of areas. The genetic algorithm (GA), which seeks to solve optimization problems using the methods of evolution, specifically survival of the fittest, was used to optimize both the feature subset, i.e. band subset, of hyper spectral data and SVM kernel parameters simultaneously. A special strategy was adopted to reduce computation cost caused by the high-dimensional feature vectors of hyper spectral data when the feature subset part of chromosome was designed. The GA-SVM method was realized using the ENVI/IDL language, and was then tested by applying to a HYPERION hyper spectral image. Comparison of the optimized results and the un-optimized results showed that the GA-SVM method could significantly reduce the computation cost while improving the classification accuracy. The number of bands used for classification was reduced from 198 to 13, while the classification accuracy increased from 88.81% to 92.51%. The optimized values of the two SVM kernel parameters were 95.0297 and 0.2021, respectively, which were different from the default values as used in the ENVI software. In conclusion, the proposed wrapper feature selection method GA-SVM can optimize feature subsets and SVM kernel parameters at the same time, therefore can be applied in feature selection of the hyper spectral data.

[1]  Riyaz Sikora,et al.  Framework for efficient feature selection in genetic algorithm based data mining , 2007, Eur. J. Oper. Res..

[2]  Cheng-Lung Huang,et al.  A GA-based feature selection and parameters optimizationfor support vector machines , 2006, Expert Syst. Appl..

[3]  Yi Liu,et al.  FS_SFS: A novel feature selection method for support vector machines , 2006, Pattern Recognit..

[4]  Chih-Jen Lin,et al.  A Practical Guide to Support Vector Classication , 2008 .

[5]  Paul S. Bradley,et al.  Feature Selection via Mathematical Programming , 1997, INFORMS J. Comput..

[6]  Ron Kohavi,et al.  Wrappers for Feature Subset Selection , 1997, Artif. Intell..

[7]  S. Sathiya Keerthi,et al.  Efficient tuning of SVM hyperparameters using radius/margin bound and iterative algorithms , 2002, IEEE Trans. Neural Networks.

[8]  Wang Chang-yao A Experimental Study on Imaging Spectrometer Data Feature Selection and Wheat Type Identification , 2006 .

[9]  Sungzoon Cho,et al.  GA-SVM wrapper approach for feature subset selection in keystroke dynamics identity verification , 2003, Proceedings of the International Joint Conference on Neural Networks, 2003..

[10]  Vladimir N. Vapnik,et al.  The Nature of Statistical Learning Theory , 2000, Statistics for Engineering and Information Science.

[11]  Qin Qian-qin Study of Band Selection Methods of Hyperspectral Image Based on Grouping Property , 2006 .

[12]  R. Reulke,et al.  Remote Sensing and Spatial Information Sciences , 2005 .

[13]  Godfried T. Toussaint,et al.  Note on optimal selection of independent binary-valued features for pattern recognition (Corresp.) , 1971, IEEE Trans. Inf. Theory.

[14]  J. Elashoff,et al.  On the choice of variables in classification problems with dichotomous variables. , 1967, Biometrika.

[15]  Sayan Mukherjee,et al.  Feature Selection for SVMs , 2000, NIPS.

[16]  Anil K. Jain,et al.  Dimensionality reduction using genetic algorithms , 2000, IEEE Trans. Evol. Comput..

[17]  Jiao Licheng,et al.  Automatic model selection for support vector machines using heuristic genetic algorithm , 2006 .

[18]  Corinna Cortes,et al.  Support-Vector Networks , 1995, Machine Learning.

[19]  Massimiliano Pontil,et al.  Support Vector Machines for 3D Object Recognition , 1998, IEEE Trans. Pattern Anal. Mach. Intell..

[20]  Li Xia Yang Qing-sheng Calibrating urban cellular automata using genetic algorithms , 2007 .

[21]  Thomas M. Cover,et al.  The Best Two Independent Measurements Are Not the Two Best , 1974, IEEE Trans. Syst. Man Cybern..

[22]  Jinbo Bi,et al.  Dimensionality Reduction via Sparse Support Vector Machines , 2003, J. Mach. Learn. Res..

[23]  Yang Zhe-hai,et al.  Feature Selection in Hyperspectral Classification Based on Tabu Search Algorithm , 2006 .

[24]  Yoshua Bengio,et al.  Gradient-Based Optimization of Hyperparameters , 2000, Neural Computation.

[25]  Sungzoon Cho,et al.  Ensemble based on GA wrapper feature selection , 2006, Comput. Ind. Eng..

[26]  Kezhi Mao,et al.  Feature subset selection for support vector machines through discriminative function pruning analysis , 2004, IEEE Transactions on Systems, Man, and Cybernetics, Part B (Cybernetics).