Postdoctoral Position in Machine Learning with a Focus on Reinforcement Learning and Model Predictive Control

Postdoctoral Position in Machine Learning with a Focus on Reinforcement Learning and Model Predictive Control

Uppsala University

Uppsala, Sweden

Duties

This project focuses on developing, analyzing, and applying new methods in reinforcement learning and model predictive control (MPC), both stochastic and deterministic formulations. the focus will be on developing methods for robust reinforcement learning and/or MPC formulations. We are particularly interested in settings where the cost or reward function is non-ergodic. In these situations, the expected costs/rewards do not capture the behaviour of individual rollouts, which calls for new constructions. We will develop and analyze generally applicable algorithms and models together with our collaborators. The postdoctoral researcher will be jointly supervised by Assistant Professor Dominik Baumann (Aalto University) and when it comes to the MPC aspects, we will collaborate with Assistant Professor Johannes Köhler at Imperial College London.

Technical building blocks may include reinforcement learning, state-space models, MPC, deep learning, and probabilistic modeling in general.

The position may include teaching up to 20% depending on availability and interest. You are expected to be able to teach in Swedish or English.

Requirements

PhD degree in machine learning, control, signal processing, or another nearby relevant area or a foreign degree equivalent to a PhD degree in machine learning, control, signal processing. The degree needs to be obtained by the time of the decision of employment. Priority will be given to applicants who have completed their degree no more than three years before the deadline for applications. Due to special circumstances, the degree may have been obtained earlier. The three-year period can be extended due to circumstances such as sick leave, parental leave, duties in labour unions, etc.

We are looking for candidates with:

• a strong interest in and experience in developing new methods in reinforcement learning and/or model predictive control;
• ability to communicate technical material effectively in oral and written form (English);
• good working knowledge of programming (preferably in Python);
• personal qualities, such as a high level of creativity, accuracy and/or a structured approach to problem solving are essential.

Publications at leading conferences in machine learning and/or leading journals in control engineering are a big plus.

Additional qualifications

Experience of working with companies is good. As a person, you are creative, meticulous and have a structured approach. When selecting applicants, we will assess their ability to independently drive their work forward, to collaborate with others, to have a professional approach and to analyze and work with complex problems. Great importance will be placed on personal qualities and personal suitability.