Neuro-symbolic AI covers methods and architectures that combine neural networks and machine learning with classical symbolic reasoning in e.g. logic. At Chalmers we apply neuro-symbolic methods in several domains, for instance:
AI for mathematics: neural networks can be used to for example suggest strategies, proof steps or extra lemmas to be used in mathematical proofs. To make sure the suggestions are valid, this can be checked by a proof assistant, a program designed to check each step and all the details in a proof.
AI for cognitive science: Can we model how (symbolic) languages evolve via communication and collaboration? We study how agents and reinforcement learning can be used to model language evolution, for example for number systems and colour naming.
AI and natural science: How can we learn symbolic expressions (to later be used in e.g. simulations and modelling) directly from experimental data?
We currently offer the course Neuro-symbolic AI on MSc level (also available for PhD students).
Faculty members:
- Associate Professor, Data Science and AI, Computer Science and Engineering
- Head of Unit, Data Science and AI, Computer Science and Engineering
- Lecturer of the Practice, Computing Science, Computer Science and Engineering
- Researcher, Data Science and AI, Computer Science and Engineering
PhD students and Postdocs:
- Doctoral Student, Data Science and AI, Computer Science and Engineering
- Doctoral Student, Data Science and AI, Computer Science and Engineering
- Doctoral Student, Data Science and AI, Computer Science and Engineering
