Earthquake detection using Remora Based Bi-Directional Long Short-Term Memory Model

In the noisy data processing and the nursing of micro earthquakes, earthquake signal recognition and seismic phase selecting are tough problems. We introduce a global deep learning model that can pick out the phases of an earthquake at the same time. Fresh years have seen an uptick in the use of technology facilitated by artificial intelligence (AI). Several industries and uses may benefit from the increased productivity and decreased need for human labour that AI technology provide. Observational earthquake seismology relies heavily on earthquake detection. In this paper, we suggest using a Deep Learning (DL) model for automated earthquake uncovering and identification. This research proposes a fresh approach to optimising Remora by combining it with a diagnostic strategy made possible by artificial intelligence. The purpose of the suggested model is to act as a classifier for detecting the presence of earthquakes, and it does so by using a (Bi-LSTM) model. The work’s originality lies in the fact that the Remora Optimization Algorithm (ROA) may be used to adjust the hyperparameter values of the Bi-LSTM model. Several simulation assessments on benchmark datasets revealed that the ROA-Bi-LSTM algorithm outperformed state-of-the-art methods..