Quake100: A Neural Network-Based Application for Predicting Earthquakes in Peru (#924)

Authors

Quinto Huamán, Carlos

Arce Oré, Paulo Cesar

Arones Hernández, Saúl Antonio

Paredes Silvestre, Luis Alberto

Inga Rodríguez, Sebastián

Rojas Cangahuala, Gladys Madeleine

Abstract

Currently, seismic events are recurrent in Peru, causing human losses and instilling fear among the population. This is primarily due to the absence of an early warning system or prediction platform that could foresee such tectonic events. Given Peru's geological location in a tectonically active area, earthquakes pose a significant threat. The main objective of this work is to implement a mobile application based on neural networks to predict the occurrence of earthquakes within a 100-day interval. The aim is to provide relevant information for risk management and disaster preparedness. This study involved the collection and preparation of a comprehensive set of historical seismic data, incorporating features such as magnitude, depth, location, and chronological sequence of events. Preliminary results indicate that neural networks have promising potential to generate reliable predictions of seismic events in Peru. In summary, this proposal contributes to the intersection of seismology and neural networks by suggesting a method for predicting seismic events in Peru using neural networks. Despite the remaining challenges, this study offers a promising path towards strengthening early warning systems and reducing seismic risk in the region. Continuing to integrate real-time data and improving neural network models can have a significant impact on the safety and resilience of Peruvian communities against seismic events in the future.