Postdoc: Probabilistic ML for Energy Materials & PV

DTU - Technical University of Denmark; Kongens Lyngby; On-site; DKK 420,000 - 540,000; Full time

Job summary

DTU Compute in Denmark seeks a postdoctoral researcher to develop probabilistic modeling and active learning methods for next-generation solar energy materials. You will work with DTU Compute and the Thin-Film Photovoltaics Group to advance machine learning for materials discovery and experimental design.

You will build models that connect synthesis conditions to material properties, publish results, and contribute to teaching and supervision of PhD and MSc/BSc students.

Qualifications

• Strong background in probabilistic ML, incl. Gaussian processes, Bayesian nets, uncertainty quantification.
• Experience implementing ML methods in Python with scientific computing frameworks.
• Strong publication record in ML or closely related fields.
• Excellent communication and interdisciplinary collaboration.
• Formal qualification: PhD degree; knowledge of materials science or photovoltaics is a plus.

Responsibilities

• Develop probabilistic models linking synthesis conditions, material properties and photovoltaic performance.
• Research active learning and Bayesian optimization for informative experiments under constraints.
• Integrate physical and chemical knowledge into models and identify informative descriptors.
• Develop multi-objective strategies for targeting bandgaps and reducing recombination losses.
• Collaborate with the experimental postdoc and DTU Electro to design experiments and interpret results.
• Publish scientific papers and present results at ML conferences and interdisciplinary venues.
• Contribute to supervision of PhD and MSc/BSc students.

Skills

Gaussian processes

Bayesian neural networks

Uncertainty quantification

Active learning

Bayesian optimization

Experimental design

Python

PyTorch

GP libraries

Education

PhD in a relevant field

Tools

Python

PyTorch

GP frameworks

Job description

DTU Compute in Denmark seeks a postdoctoral researcher to develop probabilistic modeling and active learning methods for next-generation solar energy materials. You will work with DTU Compute and the Thin-Film Photovoltaics Group to advance machine learning for materials discovery and experimental design.

You will build models that connect synthesis conditions to material properties, publish results, and contribute to teaching and supervision of PhD and MSc/BSc students.