Cognitive AI and implicit pseudo-spline wavelets for enhanced seismic prediction

Abstract

Using data from 1900 to 2024, this study developed an innovative artificial intelligence (AI)-powered framework for predicting earthquakes in Japan. By incorporating state-of-the-art cognitive computing techniques with expert seismic assessments, the proposed algorithm addresses some of the complex challenges in earthquake prediction. The model fuses AI systems with numerical methods such as the Finite Element Method (FEM) and pseudo-spline collocation techniques to simulate seismic wave propagation in a stratified spherical Earth. This study employed cognitive computing mechanisms to categorize and analyze seismic activities using a vector-based structure that compares past seismic events with predefined classifications. Moreover, the framework integrates expert knowledge of the stress distribution in the Earth’s crust to establish a comprehensive model for seismic forecasting. This AI-driven methodology provides deeper insight into seismic wave behavior and introduces a self-improving data-centric system that could support decision-making for reducing earthquake risk.