15.05.2023–14.05.2027
Principal investigator: Leila TAGHIZADEH (E101-03-2)

Nanoelectronic devices have many real-world applications, ranging from medicine to cybersecurity. Partial differential equations, which are used to model charge transport in nanoscale devices, usually have unknown parameters that cannot be calculated or measured. The initial goal of this project, opens an external URL in a new window is to formulate Bayesian inverse problems for nanodevices to estimate the model unknown parameters. The advantage of statistical Bayesian approach to solve inverse problems is that one assumes the unknown parameter, that can be a an unknown function (dependent on the spacial variable), is uncertain and random and the Bayesian approach gives a probability distribution of the quantity of interest instead of a single value estimation. To this end, we need a prior knowledge about the unknown quantity (prior distribution), and some measurement/simulated data (data likelihood distribution) which updates our prior knowledge to estimate the posterior probability distribution of the unknown quantity. The ultimate goal of the project is to improve the optimality and reliability in the design of nanodevices by efficient Bayesian algorithms and machine learning methods for semiconductor inverse problems.

Team member: Aida MOUSAVI