Skip to main navigation Skip to search Skip to main content

Two-Step Volcano Seismic Event Classification Using Metaheuristic Feature Selection and Machine Learning

  • Universidad San Francisco de Quito

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Early detection of volcanic seismic events is critical for disaster mitigation. In this paper, we propose a twostep classification method that combines metaheuristic feature selection and machine learning. Using the SeisBenchV1 dataset (75 % - 25% split with stratified k-fold), a genetic algorithm with chi-square optimization and a neural network achieved the best performance: 0.790 F1-Score and 0.979 ROC-AUC using only 23 features. The results demonstrate the potential of the method for real-time volcanic monitoring.

Original languageEnglish
Title of host publicationC3 2025 - IEEE Colombian Caribbean Conference
EditorsYesica Beltran Gomez, Paul Sanmartin Mendoza
PublisherInstitute of Electrical and Electronics Engineers Inc.
ISBN (Electronic)9798331571429
DOIs
StatePublished - 2025
Event2025 IEEE Colombian Caribbean Conference, C3 2025 - Santa Marta, Colombia
Duration: 17 Sep 202520 Sep 2025

Publication series

NameC3 2025 - IEEE Colombian Caribbean Conference

Conference

Conference2025 IEEE Colombian Caribbean Conference, C3 2025
Country/TerritoryColombia
CitySanta Marta
Period17/09/2520/09/25

Keywords

  • Feature Selection
  • Fitness Functions
  • Metaheuristics
  • Volcanic Seismic Events

Fingerprint

Dive into the research topics of 'Two-Step Volcano Seismic Event Classification Using Metaheuristic Feature Selection and Machine Learning'. Together they form a unique fingerprint.

Cite this