PhD Position Improving Symbolic AI through Learning and Reasoning

1000scholars; Delft; Remote; EUR 42,000 - 68,000; Full time

Job summary

TU Delft's Algorithmics Group invites applications for a research position exploring how machines can learn from tiny data and think like scientists. The project combines program synthesis and automated reasoning to advance scientific discovery and robotics.

You will collaborate with PhD students and postdocs, with opportunities to co-supervise. A MSc in CS/AI and experience in automated reasoning are essential, with strong programming and writing skills expected.

Qualifications

• MSc degree in Computer Science and/or Artificial Intelligence is essential.
• Prior experience in automated reasoning is essential.
• Prior experience with symbolic machine learning approaches is essential.
• Experience applying automated reasoning or machine learning in biology, chemistry, or other natural sciences is desirable.
• Good communication, organisation, and writing skills; good programming abilities to produce usable open-source software.

Responsibilities

• Research how to teach machines to learn from tiny data and think like a scientist.
• Investigate combinations of program synthesis and automated reasoning for scientific discovery and robotics.
• Collaborate with PhD students and postdocs in the Algorithmics Group and EEMCS; may co-supervise students.
• Under the supervision of Prof. Sebastijan Dumancic, contribute to high-quality open-source software for the project.

Skills

Automated reasoning

Symbolic machine learning

Programming skills

Communication skills

Organisational skills

Scientific writing

Education

MSc in Computer Science or AI

Job description

Job description

Join TU Delft in developing AI that thinks like a scientist

Job description

While all the headlines are full of AI learning from massivve datasets, in this position you will investigate how can we teach machines to learn from tiny amounts of data. You will investigate how how can we achieve this goal by making AI think like a scientist -- reason about which models and plausible and should be explored. We will build this new approach upon the combination of program synthesis and automated reasoning, and apply it to actual problems from scientific discovery and robotics.

You will join the Algorithmics Group in the Department of Software Technology, Faculty of Electrical Engineering, Mathematics and Computer Science (EEMCS). You will collaborate with PhD students and postdocs working on program synthesis and automated reasoning, and will have opportunities to co-supervise students. You will be supervised by Prof. Sebastijan Dumancic.

Job requirements

The ideal candidate will be an ambitious, proactive, reflective, and results-driven researcher. We are looking for the following:
MSc degree in Computer Science and/or Artificial Intelligence (essential) prior experience in automated reasoning (essential) prior experinece with symbolic machine learning approaches (essential) prior experience in applying automated reasoning or machine learning in biology, chemistry, or other naturla sciences (desirable) good communication, organisation, and writing skills good programming skills and desire to produce usable open-source software

TU Delft (Delft University of Technology)