Abstract
Hyperspectral data is usually high dimensional, and there is often a scarcity of available ground truth pixels . Thus the task of applying even a simple classifier such as the Gaussian Maximum Likelihood (GML) classifier usually forces the analyst to reduce the complexity of the implicit parameter estimation task. For decades, the common perception in the literature has been that the solution to this has been to reduce data dimensionality. However, as can be seen from a result by Cover [1], reducing dimensionality increases the risk of making the classification problem more complex.Using the simple GML classifier we compare state of the art dimensionality reduction strategies with a recently proposed strategy for sparsing of parameter estimates in full dimension [2]. Results show that reducing parameter estimation complexity by fitting sparse models in full dimension have a slight edge on the common approaches.
Chapter PDF
Similar content being viewed by others
Keywords
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.
References
Cover, T.M.: Geometrical and statistical properties of systems of linear inequalities with applications in pattern recognition. IEEE Transactions on Electronic Computers 14(3), 326–334 (1965)
Berge, A., Jensen, A.C., Solberg, A.S.: Sparse inverse covariance estimates for hyperspectral image classification. IEEE Trans. Geosci. Remote Sensing, Accepted for publication (2007)
Smith, M., Kohn, R.: Parsimonius covariance matrix estimation for longitudinal data. Journal of the American Statistical Association 97(460), 1141–1153 (2002)
Pouhramadi, M.: Foundations of Time Series Analysis and Prediction Theory. Wiley, Chichester (2001)
Lee, C., Landgrebe, D.: Feature extraction based on desicion boundaries. IEEE Trans. Pattern Anal. Machine Intell. 15(15), 388–400 (1993)
Kuo, B.C., Landgrebe, D.: A robust classification procedure based on mixture classifiers and nonparametric weighted feature extraction. Remote Sensing 40(11), 2486–2494 (2002)
Gamba, P.: A collection of data for urban area characterization. In: Proc. IEEE Geoscience and Remote Sensing Symposium (IGARSS’04) (2004)
Ham, J., Chen, Y., Crawford, M.M., Ghosh, J.: Investiagtion of the random forest framework for classification of hyperspectral data. Remote Sensing 43(3), 492–501 (2005)
Author information
Authors and Affiliations
Editor information
Rights and permissions
Copyright information
© 2007 Springer Berlin Heidelberg
About this paper
Cite this paper
Berge, A., Schistad Solberg, A. (2007). Improving Hyperspectral Classifiers: The Difference Between Reducing Data Dimensionality and Reducing Classifier Parameter Complexity. In: Ersbøll, B.K., Pedersen, K.S. (eds) Image Analysis. SCIA 2007. Lecture Notes in Computer Science, vol 4522. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-73040-8_30
Download citation
DOI: https://doi.org/10.1007/978-3-540-73040-8_30
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-73039-2
Online ISBN: 978-3-540-73040-8
eBook Packages: Computer ScienceComputer Science (R0)