A Data Science and Machine Learning Approach to Complex Water Data
Dr W Pang
Dr J Geris
Prof C Soulsby
No more applications being accepted
Competition Funded PhD Project (European/UK Students Only)
The wide use of sensors and telemetry in water resource systems is increasingly generating large amounts of high-resolution and complex data, the so-called Big Water Data. There is a growing interest in analysing these data in real time, especially in time-critical systems.
Recent advances in data science and machine learning have provided us with effective tools and methods for gaining more insights into such Big Water Data. For instance, the use of data clustering and anomaly detection tools can help detect anomalous spatiotemporal patterns from streaming and heterogeneous water data.
In this project we will build on our previous project IADES (Immune-inspired Anomaly Detection in Environmental Sensor Data) and further develop more effective data mining tools coupled with hydrological modelling to facilitate a better understanding of water data. These tools will help deliver early warning information, better inform the public, and trigger effective response mechanisms to deal with/prevent potential problems, such as floods and droughts.
In particular, we will explore the following topics:
(1) Adapting immune-inspired anomaly detection algorithms for real-time water data monitoring. More specifically, we are interested in detecting early signs of floods (developed over hours or days) from high-resolution rainfall and river level data (recorded every 15 minutes).
(2) Developing other bio-inspired data mining tools informed by hydrological modelling to further improve the detection and prediction performance.
(3) Exploring more data science tools for streaming water data analysis, including evolutionary clustering, matrix and tensor-based methods, and investigate their application potential in hydrological scenarios.
(4) Engaging with potential partners (SEPA, Environment Agency) for the deployment and evaluation of the developed data mining tools (e.g. as part of their flood warning and response systems).
This project may involve traveling to potential partners for engagement and research dissemination.
The successful candidate should have, or expect to have, an Honours Degree at First Class (or equivalent) in Computing Science or related disciplines
Essential: data mining and machine learning basics; programming experience in Java, Python, Ruby, or Scala.
Desirable: Sensor Data Analysis, Hydrological Data Analysis, Hydrological Modelling
This project may be funded by an EPSRC DTP. The funding will cover tuition fees and a stipend for 3 years (£14,296 p.a. in 2016/17). Due to funding restrictions the studentship is open only to UK and EU nationals with 3 years residency in the UK
This project is advertised in relation to the research areas of the disciplines of Computing Science and Geography.
Formal applications can be completed online: http://www.abdn.ac.uk/postgraduate/apply. You should apply for PhD in Computing Science, to ensure that your application is passed to the correct College for processing.
Informal inquiries can be made to Dr W Pang ([email protected]) with a copy of your curriculum vitae and cover letter. All general enquiries should be directed to the Graduate School Admissions Unit ([email protected]).