Are you interested in developing state-of-the-art statistical and machine learning techniques? Do you want your methods to be applicable to real-world complex data sets? Apply for this PhD position! Your job
This PhD position is part of the project "Getting the best predictions from complex data sets: Balancing scalability and interpretability." In this data-driven era, high-dimensional data sets hold a wealth of untapped insights. The main challenge is to optimise predictive models from these complex data sets while preserving interpretability, a crucial aspect for subsequent application in fields such as behavioural prediction and medical diagnostics. This project aims to bridge the gap between the scalability of predictive modeling techniques and their interpretability by combining rigorous statistical methods with innovative machine learning approaches.
In this project, we will optimise Bayesian penalised regression methods to provide optimal predictions in high-dimensional data sets. In addition to these more traditional statistical methods, we will explore the use of learning techniques like Generative Adversarial Networks (GANs) and Adversarial Random Forests (ARFs) in complex real-world data sets. To enhance both predictive accuracy and interpretability, we will investigate how prior knowledge can best be incorporated in these methods. Finally, the developed methods will be extensively evaluated and compared, they will be applied to real-world complex data sets from the social and medical sciences, and they will be implemented in a user-friendly R-package.
You responsibilities will be the following:
- conducting the research (literature research, developing the methods, conducting simulation studies, analysing existing data sets, preparing and storing data packages for sharing);
- writing international scientific publications and a dissertation;
- giving presentations at (inter)national scientific conferences;
- active participation in the department of Methodology and Statistics (M&S);
- knowledge utilisation: collaborating and sharing findings with applied researchers and developing user-friendly software (in the form of an R-package);
- following courses/trainings (e.g., in the context of IOPS and the local graduate school).
Your work will also include between 10% and 20% teaching tasks. You will be well guided and supported by two daily supervisors as well as a senior supervisor from M&S.