Machine learning algorithms applied to engineering seismology and earthquake engineering

Chanda, Sarit and Somala, S N (2021) Machine learning algorithms applied to engineering seismology and earthquake engineering. PhD thesis, Indian Institute of Technology Hyderabad..

Text
ce16resch11006.pdf
Restricted to Registered users only until 20 June 2024.
Download (12MB) | Request a copy

Abstract

Machine learning algorithms are used in this thesis to predict earthquake parameters for simulated and recorded events. First, the seismological engineering problem of location and magnitude is addressed using machine learning algorithms on synthetic seismograms. Only a single station/ Single component seismogram is used to estimate the earthquake location and the magnitude. The data-driven machine learning techniques are useful to extract the information from the seismograms. The Ground Motion Prediction Equations (GMPEs) like machine learning models have been developed for the different seismically active regions. Ground motion parameters such as intensity, peak ground motion, and earthquake duration are addressed using recorded data for various tectonic regimes worldwide. Ensemble learning algorithms based on majority voting have also been tried out in the thesis and for each seismically active region considered, the best performing algorithm is identified.

[error in script]

IITH Creators: