Data efficient AI refers to the ability of AI systems to learn and make accurate predictions with minimal amounts of training data. For example, in the medical field it may be challenging to obtain large amounts of labelled data due to privacy concerns. When developing new engines for planes or dimensioning windmill turbines the number of training data or tests can be limited due to high test or simulation costs. IDLab has over 20 years of experience in designing surrogate models, building a digital twin of very complex processes. It applies its expertise often within the context of engineering, facilitating the sequential Design of Experiments (DoE) and the design of virtual twins. Furthermore, IDLab also designs AI systems where you must cope with multi-accuracy data, noise, cheap-expensive constraints and uncertainty quantification.
IDLAB Ghent has expertise in (applied) research on data-efficient machine learning mainly within the domains of Electronic Design (e.g., analog design (microwave/RF)) and Engineering Design. Engineering design application domains include Electronics , Photonics, 3D printing, Automotive, Aerospace, Construction and Metallurgy.
