Unsupervised Feature Selection for Pattern Discovery in Seismic Wavefields

Andreas Köhler, Matthias Ohrnberger, Carsten Riggelsen, Frank Scherbaum
Proceedings of the Workshop on New Challenges for Feature Selection in Data Mining and Knowledge Discovery at ECML/PKDD 2008, PMLR 4:106-121, 2008.

Abstract

This study presents an unsupervised feature selection approach for the discovery of significant patterns in seismic wavefields. We iteratively reduce the number of features generated from seismic time series by first considering significance of individual features. Significance testing is done by assessing the randomness of the time series with the Wald-Wolfowitz runs test and by comparing observed and theoretical variability of features. In a second step the in-between feature dependencies are assessed based on correlation hunting in feature subsets using Self-Organizing Maps (SOMs). We show the improved discriminative power of our procedure compared to manually selected feature subsets by cross-validation applied to synthetic seismic wavefield data. Furthermore, we apply the method to real-world data with the aim to define suitable features for earthquake detection and seismic phase classification in seismic recordings.

Cite this Paper

BibTeX
@InProceedings{pmlr-v4-koehler08a, title = {Unsupervised Feature Selection for Pattern Discovery in Seismic Wavefields}, author = {Köhler, Andreas and Ohrnberger, Matthias and Riggelsen, Carsten and Scherbaum, Frank}, booktitle = {Proceedings of the Workshop on New Challenges for Feature Selection in Data Mining and Knowledge Discovery at ECML/PKDD 2008}, pages = {106--121}, year = {2008}, editor = {Saeys, Yvan and Liu, Huan and Inza, Iñaki and Wehenkel, Louis and Pee, Yves Van de}, volume = {4}, series = {Proceedings of Machine Learning Research}, address = {Antwerp, Belgium}, month = {15 Sep}, publisher = {PMLR}, pdf = {http://proceedings.mlr.press/v4/koehler08a/koehler08a.pdf}, url = {https://proceedings.mlr.press/v4/koehler08a.html}, abstract = {This study presents an unsupervised feature selection approach for the discovery of significant patterns in seismic wavefields. We iteratively reduce the number of features generated from seismic time series by first considering significance of individual features. Significance testing is done by assessing the randomness of the time series with the Wald-Wolfowitz runs test and by comparing observed and theoretical variability of features. In a second step the in-between feature dependencies are assessed based on correlation hunting in feature subsets using Self-Organizing Maps (SOMs). We show the improved discriminative power of our procedure compared to manually selected feature subsets by cross-validation applied to synthetic seismic wavefield data. Furthermore, we apply the method to real-world data with the aim to define suitable features for earthquake detection and seismic phase classification in seismic recordings.} }
Endnote
%0 Conference Paper %T Unsupervised Feature Selection for Pattern Discovery in Seismic Wavefields %A Andreas Köhler %A Matthias Ohrnberger %A Carsten Riggelsen %A Frank Scherbaum %B Proceedings of the Workshop on New Challenges for Feature Selection in Data Mining and Knowledge Discovery at ECML/PKDD 2008 %C Proceedings of Machine Learning Research %D 2008 %E Yvan Saeys %E Huan Liu %E Iñaki Inza %E Louis Wehenkel %E Yves Van de Pee %F pmlr-v4-koehler08a %I PMLR %P 106--121 %U https://proceedings.mlr.press/v4/koehler08a.html %V 4 %X This study presents an unsupervised feature selection approach for the discovery of significant patterns in seismic wavefields. We iteratively reduce the number of features generated from seismic time series by first considering significance of individual features. Significance testing is done by assessing the randomness of the time series with the Wald-Wolfowitz runs test and by comparing observed and theoretical variability of features. In a second step the in-between feature dependencies are assessed based on correlation hunting in feature subsets using Self-Organizing Maps (SOMs). We show the improved discriminative power of our procedure compared to manually selected feature subsets by cross-validation applied to synthetic seismic wavefield data. Furthermore, we apply the method to real-world data with the aim to define suitable features for earthquake detection and seismic phase classification in seismic recordings.
RIS
TY - CPAPER TI - Unsupervised Feature Selection for Pattern Discovery in Seismic Wavefields AU - Andreas Köhler AU - Matthias Ohrnberger AU - Carsten Riggelsen AU - Frank Scherbaum BT - Proceedings of the Workshop on New Challenges for Feature Selection in Data Mining and Knowledge Discovery at ECML/PKDD 2008 DA - 2008/09/11 ED - Yvan Saeys ED - Huan Liu ED - Iñaki Inza ED - Louis Wehenkel ED - Yves Van de Pee ID - pmlr-v4-koehler08a PB - PMLR DP - Proceedings of Machine Learning Research VL - 4 SP - 106 EP - 121 L1 - http://proceedings.mlr.press/v4/koehler08a/koehler08a.pdf UR - https://proceedings.mlr.press/v4/koehler08a.html AB - This study presents an unsupervised feature selection approach for the discovery of significant patterns in seismic wavefields. We iteratively reduce the number of features generated from seismic time series by first considering significance of individual features. Significance testing is done by assessing the randomness of the time series with the Wald-Wolfowitz runs test and by comparing observed and theoretical variability of features. In a second step the in-between feature dependencies are assessed based on correlation hunting in feature subsets using Self-Organizing Maps (SOMs). We show the improved discriminative power of our procedure compared to manually selected feature subsets by cross-validation applied to synthetic seismic wavefield data. Furthermore, we apply the method to real-world data with the aim to define suitable features for earthquake detection and seismic phase classification in seismic recordings. ER -
APA
Köhler, A., Ohrnberger, M., Riggelsen, C. & Scherbaum, F.. (2008). Unsupervised Feature Selection for Pattern Discovery in Seismic Wavefields. Proceedings of the Workshop on New Challenges for Feature Selection in Data Mining and Knowledge Discovery at ECML/PKDD 2008, in Proceedings of Machine Learning Research 4:106-121 Available from https://proceedings.mlr.press/v4/koehler08a.html.

Related Material