PhD in Self Adaptive Machine Learning Methods For Outcome Prediction In Real World Complex Datasets

University of Portsmouth, established in 1869 and located in the historic coastal city of Portsmouth, offers a vibrant student experience with around 28,000 students, including approximately 5,000 international learners. The university combines strong teaching and research with practical industry links and a waterfront location that enriches campus life. Its long history of technical and professional education makes it a familiar choice for students seeking applied learning in a city with excellent transport links.

Portsmouth provides a range of programs organized across ten broad subject areas, encompassing engineering, business, health professions, creative arts and sciences. Students benefit from industry placements, modern laboratories and career-focused modules that enhance employability. The university also supports research-led teaching and partnerships with local employers, giving international students opportunities to gain real-world experience alongside academic study.

Comprehensive support services help international students with visas, accommodation, language preparation and wellbeing, while student societies and campus activities foster social integration. With practical training options, placement support and access to regional industries and cultural sites, Portsmouth is well suited to students who value both academic quality and hands-on experience. Prospective applicants should investigate program specifics, placement opportunities and scholarship options before applying.

Students will join a team working on risk and mortality prediction in ICU, with access to an existing preprocessed ICU database. The project, supervised by Dr Mohamed Bader, focuses on developing self-adaptive machine learning methods for outcome prediction in real-world complex datasets.

The project will leverage techniques from High Performance Computing to handle large datasets, such as the MIMIC and eICU database which contains over 3 million ICU stays and billions of medical measurements from more than 400 hospitals. Another key area of investigation is consumer behaviour prediction for digital marketing, in collaboration with a leading digital marketing firm. This part of the project aims to develop machine learning methods capable of predicting consumer behaviour and optimising marketing strategies. Students will need strong programming skills in Python and R, as well as solid machine learning and data mining skills. Experience in deep learning, computational intelligence, data mining, machine learning, health informatics, and/or data analytics is highly desired.