University of Oslo
The PhD position is one of four PhD and post-doc positions within the UiORealArt Convergence environment at the University of Oslo. Convergence environments are interdisciplinary research groups that will aim to solve grand challenges related to health and the environment. They are funded by UiO’s interdisciplinary strategic area UiO:Life Science www.uio.no/life-science.
The aim of this Convergence environment is to improve causal inference in perinatal pharmaco-epidemiology using machine learning approaches on real-world and artificial data” (UiORealArt, 2022-2026). The project unites expertise within natural sciences and medicine (pharmacology, machine learning, bioinformatics, statistics, genetics, and epigenetics) with social/educational sciences (psychology, language development and educational attainment).
Perinatal pharmaco-epidemiology refers to a field that studies the safety of medication use during pregnancy. Causal inference in perinatal pharmaco-epidemiology is extremely challenging because of the nature of allowed data collection procedures while the methodological progress has on some aspects stagnated. There is currently a surge of interest in applying machine learning (ML) to epidemiological questions under the assumption that it would improve causal inference. There has also been an increased interest in incorporating high-dimensional molecular data together with epidemiological data from national health registries to strengthen causal inference. However, both the application of ML and the incorporation of high-dimensional molecular data appear to have been mostly based on hype and popular trends rather than a deeper appreciation for problem characteristics and suited inductive biases.
The PhD candidate will build on the combined expertise of epidemiologists and machine learners involved in RealArt and will work at the intersection of bioinformatics, machine learning and causal inference. First, while causality has traditionally been focused on linear relations involving data of limited size/dimensionality, there is currently an increasing interest in considering higher complexity (non-linear) relations and using data of larger size and higher dimensionality. Second, while machine learning has traditionally been focused on reflecting statistical associations in fixed underlying distributions – based on a dataset assumed to be an iid representation of this underlying distribution – there has lately been an increased realization that real-world application of machine learning very often involves substantial domain shifts, and that causality provides perspectives that are very useful to reason about and to improve the generalizability/robustness of ML models in face of such challenges.
The PhD student will mainly lead projects related to synthetic data generation, applied machine learning and benchmarking within the scope of RealArt. The following are provisional project directions in sequential order: 1) develop a framework for simulating high-dimensional molecular marker datasets under a defined causal structure with varying levels of correlations. 2) develop a benchmark suite to enable the assessment of the feasibility of drawing causal conclusions through the incorporation of molecular markers in perinatal epidemiology. 3) Assessment of the performance of analytical methods in recovering the ground truth associations between molecular markers and covariates in perinatal epidemiology datasets.
The Faculty of Mathematics and Natural Sciences has a strategic ambition to be among Europe’s leading communities for research, education and innovation. Candidates for this fellowship will be selected in accordance with this, and expected to be in the upper segment of their class with respect to academic credentials.
- Applicants must hold a Master’s degree equivalent to a Norwegian Master’s degree in Bioinformatics/Informatics, Data Science, Computer Science, Statistics or a related field. Candidates without a Master’s degree have until 30 June 2023 to complete the final exam.
- Strong programming skills in Python
- Excellent communication skills in English, both oral and written
Although not required, the following qualifications will be an advantage in the assessment of the applicants:
- Documented experience and/or decent grasp of machine learning or statistical methodologies
- Documented experience in handling data from high-throughput molecular data (omics)
- Projects or coursework in causal inference
- Projects or coursework involving data simulations
The norm is as follows:
- the average grade point for courses included in the Bachelor’s degree must be C or better in the Norwegian educational system
- the average grade point for courses included in the Master’s degree must be B or better in the Norwegian educational system
- the Master’s thesis must have the grade B or better in the Norwegian educational system
- Fluent oral and written communication skills in English
English requirements for applicants from outside of EU/ EEA countries and exemptions from the requirements:
The purpose of the fellowship is research training leading to the successful completion of a PhD degree.
The fellowship requires admission to the PhD programme at the Faculty of Mathematics and Natural Sciences. The application to the PhD programme must be submitted to the department no later than two months after taking up the position. For more information see:
- Salary NOK 501 200 – 544 400 per year depending on qualifications and seniority as PhD Research Fellow (position code 1017)
- Attractive welfare benefits and a generous pension agreement
- Vibrant international academic environment
- Career development programmes
- Oslo’s family-friendly surroundings with their rich opportunities for culture and outdoor activities
How to apply
The application must include:
- Cover letter: The candidates are advised to describe in the cover letter how they meet the required qualifications and if they meet any of the desirable qualifications. The cover letter should also include a statement of motivation, research interest and short summary of any prior scientific work.
- CV (summarizing education, skills, any previous positions, and other qualifying activities such as teaching experience, and administrative experience)
- Copies of educational certificates (original Bachelor and Master’s degree diplomas), academic transcript of records and letters of recommendation
- Any examples of code from the applicant’s previous work shared through a GitHub link that allows the evaluation of programming skills
- List of publications (if any) and academic work (e.g., master’s thesis or similar) that the applicant wishes to be considered by the evaluation committee
- Names and contact details of 2-3 references (name, relation to candidate, e-mail and telephone number)
The application with attachments must be delivered in our electronic recruiting system, please follow the link “apply for this job”. Foreign applicants are advised to attach an explanation of their University’s grading system. Please note that all documents should be in English (or a Scandinavian language).
In assessing the applications, special emphasis will be placed on the documented, academic qualifications, the statement of motivation and the candidate’s personal suitability. Interviews with the best-qualified candidates will be arranged.
It is expected that the successful candidate will be able to complete the project in the course of the period of employment.
synthetic data generation
applied machine learning