PhD-Student Zayan Ahsan Ali

Zayan is pursuing his PhD under the guidance of Prof.Dr. Waldmann. In his master's thesis, which is titled "Machine learning based parameterization for magnetic data of single-molecule magnets", he applied a machine learning-based approach to Single Molecule Magnets (SMMs) with a specific focus on addressing the over-parametrization challenge. His methodology involved using a combination of techniques such as Variational Autoencoders (VAEs) and Invertible Neural Networks (INNs) to set up a machine learning framework. This framework not only identifies non-physical parameterizations within the data but also builds a connection to the Hamiltonian model parameters. This is achieved through the innovative use of a "Black-Box" Model, showcasing the versatility of Machine Learning in this domain.

Currently, Zayan is working on expanding capability of the implemented framework to include higher order Hamiltonian parameters. This involves several challenges such as ideal preprocessing, parameter importance and effective mapping of high dimensional spaces.