Skip to main content
main-content

Tipp

Weitere Kapitel dieses Buchs durch Wischen aufrufen

2022 | OriginalPaper | Buchkapitel

On Training Set Selection in Spatial Deep Learning

verfasst von: Eligius M. T. Hendrix, Mercedes Paoletti, Juan Mario Haut

Erschienen in: High-Dimensional Optimization and Probability

Verlag: Springer International Publishing

share
TEILEN

Abstract

The careful design of experiments in spatial statistics aims at estimating models in an accurate way. In the field of spatial deep learning to classify spatial observations, the training set used to calibrate a model or network is usually determined in a random way in order to obtain a representative sample. This chapter will sketch with examples that this is not necessarily the best way to proceed. Moreover, as in some cases windows are used to smooth signals, overlap may occur in the spatial data. On the one hand, this implies auto-correlation in the training set and, on the other hand, a correlation among pixels used for training and for testing. Our question is how to measure such an overlap and how to steer the selection of training sets. We describe an optimization problem to model and minimize the auto-correlation. A simple example is used to capture the concepts of design of experiments versus training set selection and the measurement of the overlap.
Literatur
1.
Zurück zum Zitat G.E.P. Box, H.L. Lucas, Design of experiments in non-linear situations. Biometrika 46(1–2), 77–90 (1959) MathSciNetCrossRef G.E.P. Box, H.L. Lucas, Design of experiments in non-linear situations. Biometrika 46(1–2), 77–90 (1959) MathSciNetCrossRef
2.
Zurück zum Zitat L. de Haan, Estimation of the minimum of a function using order statistics. J. Am. Stat. Assoc. 76(374), 467–469 (1981) MathSciNetCrossRef L. de Haan, Estimation of the minimum of a function using order statistics. J. Am. Stat. Assoc. 76(374), 467–469 (1981) MathSciNetCrossRef
3.
Zurück zum Zitat M.A. Friedl, C. Woodcock, S. Gopal, D. Muchoney, A.H. Strahler, C. Barker-Schaaf, A note on procedures used for accuracy assessment in land cover maps derived from AVHRR data. Int. J. Remote Sensing 21(5), 1073–1077 (2000) CrossRef M.A. Friedl, C. Woodcock, S. Gopal, D. Muchoney, A.H. Strahler, C. Barker-Schaaf, A note on procedures used for accuracy assessment in land cover maps derived from AVHRR data. Int. J. Remote Sensing 21(5), 1073–1077 (2000) CrossRef
4.
Zurück zum Zitat L.M. Haines, Optimal design for neural networks. Lect. Notes-Monogr. Series 34, 152–162 (1998) CrossRef L.M. Haines, Optimal design for neural networks. Lect. Notes-Monogr. Series 34, 152–162 (1998) CrossRef
5.
Zurück zum Zitat R. Hänsch, A. Ley, O. Hellwich, Correct and still wrong: The relationship between sampling strategies and the estimation of the generalization error, in 2017 IEEE International Geoscience and Remote Sensing Symposium (IGARSS) pp. 3672–3675 (2017) R. Hänsch, A. Ley, O. Hellwich, Correct and still wrong: The relationship between sampling strategies and the estimation of the generalization error, in 2017 IEEE International Geoscience and Remote Sensing Symposium (IGARSS) pp. 3672–3675 (2017)
6.
Zurück zum Zitat E.M.T. Hendrix, Global Optimization At Work. Ph.D. Thesis, Wageningen Agricultural University, 1998 E.M.T. Hendrix, Global Optimization At Work. Ph.D. Thesis, Wageningen Agricultural University, 1998
7.
Zurück zum Zitat J. Lange, G. Cavallaro, M. Götz, E. Erlingsson, M. Riedel, The influence of sampling methods on pixel-wise hyperspectral image classification with 3d convolutional neural networks, in IGARSS 2018—2018 IEEE International Geoscience and Remote Sensing Symposium (2018), pp. 2087–2090 J. Lange, G. Cavallaro, M. Götz, E. Erlingsson, M. Riedel, The influence of sampling methods on pixel-wise hyperspectral image classification with 3d convolutional neural networks, in IGARSS 2018—2018 IEEE International Geoscience and Remote Sensing Symposium (2018), pp. 2087–2090
8.
Zurück zum Zitat J. Liang, J. Zhou, Y. Qian, L. Wen, X. Bai, Y. Gao, On the sampling strategy for evaluation of spectral-spatial methods in hyperspectral image classification. IEEE Trans. Geosci. Remote Sensing 55(2), 862–880 (2017) CrossRef J. Liang, J. Zhou, Y. Qian, L. Wen, X. Bai, Y. Gao, On the sampling strategy for evaluation of spectral-spatial methods in hyperspectral image classification. IEEE Trans. Geosci. Remote Sensing 55(2), 862–880 (2017) CrossRef
9.
Zurück zum Zitat W.G. Müller, Collecting Spatial Data (Springer, Berlin, 2000) W.G. Müller, Collecting Spatial Data (Springer, Berlin, 2000)
10.
Zurück zum Zitat L. Palagi, Global optimization issues in deep network regression: an overview. J. Global Optim. 73, 239–277 (2019) MathSciNetCrossRef L. Palagi, Global optimization issues in deep network regression: an overview. J. Global Optim. 73, 239–277 (2019) MathSciNetCrossRef
Metadaten
Titel
On Training Set Selection in Spatial Deep Learning
verfasst von
Eligius M. T. Hendrix
Mercedes Paoletti
Juan Mario Haut
Copyright-Jahr
2022
DOI
https://doi.org/10.1007/978-3-031-00832-0_9

Stellenausschreibungen

Anzeige

Premium Partner