Short-term Earthquake Prediction via Recurrent Neural Network Models: Comparison among vanilla RNN, LSTM and Bi-LSTM

Earthquake prediction has raised many concerns nowadays, due to the massive loss caused by earthquakes, as well as the significance of accurate forecasting. Lots of trials have been investigated and experimented but few achieved satisfying results on short-term prediction (i.e., usually those earthquakes that will happen in three months). It is...

From in-situ measurement to regression and time series models: An overview of trends and prospects for building performance modelling

Data analysis methodologies are crucial to learn insights from data and to create more trust in the assumptions used for energy performance assessment. Indeed, continuous performance monitoring should become a more diffuse practice in order to improve our design and operation strategies for the future. This is an essential step to reduce...

Multivariate Methods for Coastal and Offshore Risks

This thesis investigates how selected multivariate probabilistic methods can be adapted for risk analysis and decision making in coastal and offshore environments. In particular, the thesis makes a contribution to decision support tools for risk reduction efforts in coastal environments and to statistical simulation methods for wave conditions....

Monitoring civil infrastructure using satellite radar interferometry

Satellite radar interferometry (InSAR) is a precise and efficient technique to monitor deformation on Earth with millimeter precision. Most InSAR applications focus on geophysical phenomena, such as earthquakes, volcanoes, or subsidence. Monitoring civil infrastructure with InSAR is relatively new, with potential for operational applications,...

The Quality of Lagged Products and Autoregressive Yule–Walker Models as Autocorrelation Estimates

The sample autocorrelation function is defined by the mean lagged products (LPs) of random observations. It is the inverse Fourier transform of the raw periodogram. Both contain the same information, and the quality of the full-length sample autocorrelation to represent random data is as poor as that of a raw periodogram. The autoregressive (AR)...

Calibration of transient groundwater models using time series analysis and moment matching

A comprehensive and efficient approach is presented for the calibration of transient groundwater models. The approach starts with the time series analysis of the measured heads in observation wells using all active stresses as input series, which may include rainfall, evaporation, surface water levels, and pumping. The time series analysis...