elib
DLR-Header
DLR-Logo -> http://www.dlr.de
DLR Portal Home | Imprint | Privacy Policy | Contact | Deutsch
Fontsize: [-] Text [+]

On the Effect of Spatially Non-disjoint Training and Test Samples on Estimated Model Generalization Capabilities in Supervised Classification with Spatial Features

Geiß, Christian and Aravena Pelizari, Patrick and Schrade, Henrik and Brenning, Alexander and Taubenböck, Hannes (2017) On the Effect of Spatially Non-disjoint Training and Test Samples on Estimated Model Generalization Capabilities in Supervised Classification with Spatial Features. IEEE Geoscience and Remote Sensing Letters, 14 (11), pp. 2008-2012. IEEE - Institute of Electrical and Electronics Engineers. ISSN 1545-598X

Full text not available from this repository.

Official URL: http://ieeexplore.ieee.org/document/8046029/

Abstract

In this letter, we establish two sampling schemes to select training and test sets for supervised classification. We do this in order to investigate whether estimated generalization capabilities of learned models can be positively biased from the use of spatial features. Numerous spatial features impose homogeneity constraints on the image data, whereby a spatially connected set of image elements is attributed identical feature values. In addition to a frequent occurrence of intrinsic spatial autocorrelation, this leads to extrinsic spatial autocorrelation with respect to the image data. The first sampling scheme follows a spatially random partitioning into training and test sets. In contrast to that, the second strategy implements a spatially disjoint partitioning, which considers in particular topological constraints that arise from the deployment of spatial features. Experimental results are obtained from multi- and hyperspectral acquisitions over urban environments. They underline that a large share of the differences between estimated generalization capabilities obtained with the spatially disjoint and non-disjoint sampling strategies can be attributed to the use of spatial features, whereby differences increase with an increasing size of the spatial neighborhood considered for computing a spatial feature. This stresses the necessity of a proper spatial sampling scheme for model evaluation to avoid overoptimistic model assessments.

Item URL in elib:https://elib.dlr.de/115167/
Document Type:Article
Title:On the Effect of Spatially Non-disjoint Training and Test Samples on Estimated Model Generalization Capabilities in Supervised Classification with Spatial Features
Authors:
AuthorsInstitution or Email of AuthorsAuthors ORCID iD
Geiß, Christianchristian.geiss (at) dlr.deUNSPECIFIED
Aravena Pelizari, Patrickpatrick.aravenapelizari (at) dlr.deUNSPECIFIED
Schrade, Henrikhenrik.schrade (at) dlr.deUNSPECIFIED
Brenning, Alexanderalexander.brenning (at) dlr.deUNSPECIFIED
Taubenböck, Hanneshannes.taubenboeck (at) dlr.deUNSPECIFIED
Date:November 2017
Journal or Publication Title:IEEE Geoscience and Remote Sensing Letters
Refereed publication:Yes
Open Access:No
Gold Open Access:No
In SCOPUS:Yes
In ISI Web of Science:Yes
Volume:14
Page Range:pp. 2008-2012
Publisher:IEEE - Institute of Electrical and Electronics Engineers
ISSN:1545-598X
Status:Published
Keywords:Hyperspectral images, model generalization capability, morphological profiles (MPs), multispectral images, random forests (RFs), spatial features, supervised classification
HGF - Research field:Aeronautics, Space and Transport
HGF - Program:Space
HGF - Program Themes:Earth Observation
DLR - Research area:Raumfahrt
DLR - Program:R EO - Erdbeobachtung
DLR - Research theme (Project):R - Vorhaben Sicherheitsrelevante Erdbeobachtung, R - Remote sensing and geoscience
Location: Oberpfaffenhofen
Institutes and Institutions:German Remote Sensing Data Center > Geo Risks and Civil Security
Deposited By: Geiß, Christian
Deposited On:09 Nov 2017 09:52
Last Modified:08 Mar 2018 18:31

Repository Staff Only: item control page

Browse
Search
Help & Contact
Information
electronic library is running on EPrints 3.3.12
Copyright © 2008-2017 German Aerospace Center (DLR). All rights reserved.