The International Archives of the Photogrammetry, Remote Sensing and Spatial Information Sciences. Vol. XXXVII. Part B7. Beijing 2008 
SFFS method, supervised PSBS yielded 0.04 percent and 0.72 
percent improvement on OA and AA, respectively. In the 
conventional BS methods, the feature size increases by adding a 
new feature to the last feature set based on maximizing 
discriminant criteria or accuracy measures. In contrast, in the 
PSBS method the K-means is computed individually for the 
new feature size. As a result, each new feature size is 
independent from last feature set and has its own discrimination 
potential. 
Another merit of this method lies in the knowledge based 
dimensionality reduction based on the spectral library. As a 
result, this method allows users to employ limited sample size 
and to treat hyperspectral data like multi-spectral data in 
information extraction tasks. 
REFERENCES 
References from Journals: 
Kudo, M. and Sklansky, J., “Comparison of algorithms that 
select features for pattern classifiers,” Pattern Recognition, vol. 
33, no. 1, pp. 25-41, Jan. 2000. 
Mao, K.Z, “Feature subset selection for support vector 
machines through discriminative function pruning analysis” 
IEEE TRANSACTIONS ON SYSTEMS, MAN, AND 
CYBERNETICS—PART B: CYBERNETICS, vol.34, no.l, 
pp.60-67, Feb. 2004. 
Martinez-Uso,A., Pla, F., Sotoca, J. and Garcia- 
Sevilla, P. “Comparison of Unsupervised Band Selection 
Methods for Hyperspectral Imaging," Proc. Iberian Conference 
on Pattern Recognition and Image Analysis (IbPRIA07), vol. I 
(Springer LNCS 4477), pp. 30-38, Girona, Spain, June 2007. 
Mart'mez-Us'o,A., Pla,F., Sotoca,J.M. P., ’’Clustering-based 
multispectral band selection using Garc'ia-Sevilla,mutual 
information”The 18th International Conference on Pattern 
Recognition (ICPR'06) IEEE computer society 2006. 
Mitra, P., murthy, C.A. pal,s.k “Unsupervised feature selection 
using feature similarity” IEEE transactions on pattern analysis 
and machine intelligence , vol.24, No.3, march 2002. 
Pal, M. “Support vector machine based feature selection for 
land cover classification: a case study with DAIS hyperspectral 
data”, International Journal of Remote Sensing, Volume 27, pp. 
2877-2894 Number 14, 20 July 2006 . 
Verzakov, S., Paclik,P., Duin,R.P.W, “Feature Shaving for 
Spectroscopic Data”. Lecture Notes in Computer Science, 
Springer-Verlag, Vol. 3138 Berlin Heidelberg New York ,pp. 
1026-1033. .2004. 
References from Books: 
Guyon,I., Gunn, S. , Nikravesh, M. , Zadeh, A., 
Extraction : Foundations and Applications 
3540354875,Springer; 1 edition ,2006. 
Varshney,P.K., Arora,M.K. ’’Advanced Image Processing 
Techniques for Remotely Sensed Hyperspectral Data” Springer, 
Berlin, 2004. 
References from Other Literature: 
Pekalska,E./’dissimilarity representations 
recognition” Phd thesis, Deleft University,2005. 
451