QML for reduced order density matrix time propagation

pemmaras commented 2 years ago

Description

Reduced density matrices (RDMs) play an important role in the theory of open quantum systems and functional approaches that aim to circumvent the exponential scaling associated with the full Hilbert space when approximating observable properties. This project aims to explore quantum machine learning (QML) of time propagators for evolving subsystem densities or density matrices of lattice models comprised of spin or Fermionic degrees of freedom. Our goal is to assess the architectural and data requirements associated with Quantum Neural Networks (QNNs) for accurately predicting the future non-Markovian time evolution of RDMs based on past histories. Our project aims to reproduce and extend recent research in this area: https://arxiv.org/abs/2201.11647 and will be conducted in collaboration with the Sanvito group at Trinity College Dublin, Ireland.

Deliverables

The primary goal of this work is to conduct QML research into predicting the time evolution of quantum subsystems which will lead to a journal publication. Additionally we hope to contribute QML models we develop to the quantum open source community in the form of Qiskit tutorials

Mentors details

Mentor 1
- Name: Das Pemmaraju
- GitHub ID: @pemmaras
- What they do: Computational Physicist and Qiskit Advocate

Number of mentees

1

Type of mentees

Mentee 1
- Required:
- Graduate level Quantum Mechanics coursework
- Nice to have:
- Qiskit coding experience
- QML basics

hykavitha commented 2 years ago

@pemmaras @HuangJunye : Hey both, This fits my area. At my day job I do QML, Would like to contribute to this work.

ichen17 commented 2 years ago

It's interesting for me. I have some knowledge of QML and also the quantum open system. I also like to contribute to this project.

pemmaras commented 2 years ago

Hi @hykavitha, @ichen17 Thank you for your interest! Are you applying to the QAMP fall 22 program? If so, please indicate your interest on the application form. If you are not applying via the QAMP program there might also be an opportunity to collaborate. Reach out to me (@Das Pemmaraju) over Qiskit slack and we can discuss options.

hykavitha commented 2 years ago

@pemmaras : I'm working on issue number #35 so will drop from this.