Physics-informed neural networks for efficient encoding of ab-initio DFT density matrices for quantum computers.

Submitted by obm on

Physics-informed neural networks for efficient encoding of ab-initio DFT density matrices for quantum computers.

Physics-informed neural networks embed the knowledge of physical laws in the learning process

You'll need to implement a physics-informed autoencoder that transcribes a density functional obtained from an ab-initio calculation into a form suitable for quantum computation. The initial appointment will be for one year with a possible extension to two years.

Effective dimension of a Quantum Neural network increases the performance of a model facing new information.

Position requirements

Must haves:

  •  A degree from a closely related field(s)
  •  Experience with computational techniques.
  • Good communication skills in English.

 Considered as an advantage:

  • Experience with Quantum computers
  • Experience with Neural Networks and Artificial Intelligence
  • Background in code development

The applicants should include their CV and a brief motivation letter along with at least two letters of reference, to be sent directly to mbaris@metu.edu.tr.

More material

  1. "Quantum computational advantage with a programmable photonic processor" https://www.nature.com/articles/s41586-022-04725-x
  2. https://qiskit.org/documentation/machine-learning/tutorials/01_neural_networks.html
  3. Fundamentals of Time-Dependent Density Functional Theory Miguel A.L. Marques, Neepa T. Maitra, Fernando M.S. Nogueira, E.K.U. Gross, Angel Rubio LNP, volume 837 Springer
  4. https://xanadu.ai/
  5. https://qiskit.org/textbook/preface.html