lyn1874/fedsers

Federated learning on the simulated surface-enhanced Raman spectra

This repository provides the implementation of FedAvg on top of the simulated surface-enhanced Raman spectra.

Requirement

git clone https://github.com/lyn1874/fedsers.git
cd sers_fl
conda env create -f sers_fl.yaml
conda activate torch_dl

Data preparation

We generate the SERS maps using the function save_data() from the file data.prepare_sers_data.py. Basically, different types of SERS maps can be generated using the function save_data(). For example, pure SERS maps, SERS maps with contaminants at a fixed Raman wavenumber, and SERS maps with contaminants at different Raman wavenumber. In this experiment, we simply use the SERS maps without any contaminants.

Training

We train fedavg using the following command

sbatch brun.sh

Result

Citation

If you use the data generator, please cite:

@Article{D3AN00446E,
author ="Li, Bo and Zappalá, Giulia and Dumont, Elodie and Boisen, Anja and Rindzevicius, Tomas and Schmidt, Mikkel N. and Alstrøm, Tommy S.",
title  ="Nitroaromatic explosives’ detection and quantification using an attention-based transformer on surface-enhanced Raman spectroscopy maps",
journal  ="Analyst",
year  ="2023",
volume  ="148",
issue  ="19",
pages  ="4787-4798",
publisher  ="The Royal Society of Chemistry",
doi  ="10.1039/D3AN00446E",
url  ="http://dx.doi.org/10.1039/D3AN00446E",
}

lyn1874 / fedsers