arXiv:2606.10069v1 Announce Type: new Abstract: In this paper we build upon a previous study in which we demonstrated, using XGBoost and earthquake catalogue data from Japan and Chile, that a set of 60 seismic statistical features (SSFs) had much greater predictive value than a set of 428 generic time series features from the tsfresh package. We here extend this previous work in two key ways, focusing on data from Japan as a large dataset is necessary in order to allow for the training of a deep learning (autoencoder) model. First, we move from whole-region prediction (considering, for each ca
This research builds on previous work using machine learning for seismic assessment, indicating a continuous advancement in applying AI to natural disaster prediction.
Improved seismic hazard assessment can lead to better early warning systems and infrastructure planning, potentially saving lives and reducing economic damage in earthquake-prone regions.
This research suggests a more refined and potentially accurate method for predicting seismic hazards using deep learning, moving beyond simpler statistical approaches.
- · AI researchers
- · Geophysical science
- · Disaster preparedness agencies
- · Insurance industry
- · Regions without advanced AI infrastructure
- · Traditional seismic analysis methods
More accurate localized seismic hazard maps can be developed.
Urban planning and building codes in earthquake zones could be proactively updated based on higher fidelity risk assessments.
The application of advanced AI in geophysical monitoring could accelerate, leading to similar breakthroughs in other natural hazard predictions.
This signal links to a primary source. Continuum Brief monitors and indexes it as part of the live intelligence stream — we do not republish source content.
Read at arXiv cs.LG