David Eklund

David is a researcher in applied artificial intelligence.

David holds a PhD in mathematics and a Master's degree in engineering physics from the Royal Institute of Technology. Davids research experience spans applied algebra, geometry and machine learning. He also has industry experience in the field of formal verification and railway signalling. Lately, David has worked in AI related projects in process chemistry, cybersecurity, privacy and formal verification.

More about Davids research:

Formal verification of AI-systems: Artificial Intelligence is increasingly applied in safety-critical domains such as autonomous vehicles, energy systems and the medical sector. This will lead to an increased demand for testing and verification of AI-systems which in part involves the use of formal methods to prove or disprove requirements that the models should fulfill. Requirements can be related to safety or the functionality of the model.

Hybrid modelling: machine learning is combined with physical models. This may be necessary when the mechanisms governing the system under modelling is only partially known, which is common for example in process chemisty. A hybrid model can be trained on less data than a purely statistical model since known properties such as physical conservation laws are built into the model.

Federated machine learning: to preserve data privacy, machine learning models can be trained locally without the sharing of private data. The local models are combined through distributed machine learning to a global model. This means that individuals or organisations can share knowledge without sharing sensitive data. To ensure that private information cannot be extracted from the local models transmitted over the network, these models may be encrypted homomorphically before transfer.