Project

Data Efficient Machine Learning for Engineering Applications

Duration
01 October 2020 → Ongoing
Funding
Regional and community funding: Special Research Fund
Research disciplines
  • Engineering and technology
    • High performance computing
    • Modelling and simulation
    • Numerical computation
Keywords
Simulations Data Efficiency Engineering Applications Surrogate Models Machine Learning
 
Project description

The goal of this research is the development of novel techniques in surrogate modelling. Surrogate models are fast approximative models for otherwise complex and time-consuming high-fidelity computer simulations. They can be used in diverse engineering applications (electronics, mechanical engineering, etc.) for optimisation, exploration of the design space, sensitivity analysis, etc., where the computation of additional computer simulations would be too costly.

The construction of surrogate models requires the use of machine learning techniques on the outcome of computer simulations. Data efficiency is essential to extract the maximal amount of useful information from the smallest possible set of data points, as each new data point requires an additional costly high-fidelity simulation.