Mobility Data Driven Privacy-preserving Model For COVID-19 Case Detection

This is the Pytorch implementation of the privacy-preserving model for COVID-19 case detection.

Requirements

• torch>=1.7.0
• numpy>=1.23.4
• scikit-learn>=1.1.3
• tensorboard>=2.11.0
• tqdm>=4.64.1
• pyyaml=6.0
• scipy>=1.9.3

Dependency can be installed with the following command:

pip install -r requirements.txt

Data Preparation

For the consideration of the user privacy, and avoid malicious usage of mobility data,the original mobility data is not publicly released. We will instead make the dataset available upon request to the corresponding author to researchers in the field for scientific purposes.

The preprocessed files (graph construction files, health status labels, etc.) for SARS-CoV-2 dataset and Omicron dataset, are available at dataset.tar.gz, and should be decompressed and put into the folder ./dataset/ with the following command:

tar -xvzf dataset.tar.gz

Model Training

Here are commands for training the model on both SARS-CoV-2 dataset and Omicron dataset.

python train.py

To train on the different datset, please manually modify the "dataset" item in the config file .config/config.yaml.

• SARS-CoV-2 Dataset

      #Training
      model: HGNN_time
      dataset: primitive
      fts_type: time
      max_epoch: 1000
      n_hid: 672
      lr: 0.001
      milestones: [100, 200, 300]
      gamma: 0.7
      drop_out: 0.2
      print_freq: 10
      weight_decay: 0
      decay_step: 100
      decay_rate: 0.7
      early_stopping: 20

• Omicron Dataset

      #Training
      model: HGNN_time
      dataset: omicron
      fts_type: time
      max_epoch: 1000
      n_hid: 672
      lr: 0.001
      milestones: [100, 200, 300]
      gamma: 0.7
      drop_out: 0.2
      print_freq: 10
      weight_decay: 0
      decay_step: 100
      decay_rate: 0.7
      early_stopping: 20