Automatic First Arrival Picking on P-Wave Seismic Signal Using Support Vector Machine Method
- 1 June 2020
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 1-6
Abstract
Automatic First Arrival Picking is a system that can get a P-Wave or the first wave that comes in an earthquake wave. Because of the P-Wave is the first wave to come, it needs research that can get the arrival of P-Wave automatically. The aim of this study is to create an Automatic First Arrival Picking system and to test the performance of methods that will later get P-Wave Picking results and also to get the accuracy of the Support Vector Machine (SVM) as its classification method. First, earthquake sample data must go through the Feature Extraction stage so that the feature results can be used as input to the SVM classification method. In this study sample data S-Wave and Noise are considered as No P-Wave, so there are only two classifications in SVM, namely P-Wave and No P-Wave. The results of this research were got an Automatic First Arrival Picking system with an accuracy performance of 88.00%, precision of 90.00%, recall of 73.50%,f1-score of 78.00% with certain time windowing, data partition, and regularization (C) parameter.Keywords
This publication has 3 references indexed in Scilit:
- Generalized Seismic Phase Detection with Deep LearningBulletin of the Seismological Society of America, 2018
- Toward Real-Time Seismic Feature Analysis for Bright Spot Detection: A Distributed ApproachIEEE Journal of Selected Topics in Applied Earth Observations and Remote Sensing, 2017
- An AutomaticP‐Phase Arrival‐Time PickerBulletin of the Seismological Society of America, 2016