PhD - Detecting Anomalies from Satellite and Terrestrial Data Sources by Data Analytics
Date added: 08 March 2019
Organisation: FindAPhD on behalf of Ulster University
Location: Ulster, UK
Applications are invited for a European Space Agency (ESA) funded PhD studentship tenable in the Faculty of Computing, Engineering and the Built environment at the Jordanstown Campus.
Detecting anomalies from data sequences obtained from the space and the torrential data sources available is to require discovering change points within the data sequences. These change points will be formulated to be normal and abnormal changes, which are in turn referred to as anomalies. In our previous studies, three methods -wavelet, statistical martingales and fuzzy-inspired - have been explored, and the respective algorithms have been developed. These results will underpin a new strategy of detecting change points within data sequences and further development of new detection algorithms.
This project aims to develop more sophisticated change detection analytics by identifying change points and comparing two distributions: one estimated from the recorded data sequences and another from dynamic streams. This development may also involve application of the current state of arts deep learning technology. The algorithms developed will be employed to analyse electromagnetic data recorded by the Swarm satellites, China Seismo Electromagnetic Satellite (CSES), CSELF network, etc., with a focus on investigating the correlation between abnormal changes and natural hazards, such as earthquakes. The analysis results will be used to develop an intelligent system for risk assessment of natural hazards and forecast.
Applications accepted all year round until the position is filled.
Learn more about the opportunity and how to apply.