Pre-trained Online Contrastive Learning for Insurance Fraud Detection

Usage

Setup your env

Run the following command to set up the environment:

conda env create -f environment.yaml

Preprocess the data

Run the following command to prepare for the data.

conda activate pygvenv
python preprocess.py

Run the code

Run the following command to run POCL

python main.py

Project Structure

• main.py: The file that contains the training process of POCL will input the results of the entire process and plot the images.
• preprocess.py: The file contains code for modeling a healthcare insurance dataset from tabular data into a temporal graph data structure.
• models.py: The file contains the main model code for POCL.
• tools.py: The file contains some helper functions.

Citing

If you find POCL is useful for your research, please consider citing the following papers:

@inproceedings{POCL,
    title={Pre-trained Online Contrastive Learning for Insurance Fraud Detection},
    author={Zhang, Rui and Cheng, Dawei and Yang, Jie and Ouyang, Yi and Wu, Xian and Zheng, Yefeng and Jiang, Changjun},
    booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
    volume={38},
    number={20},
    pages={22511--22519},
    year={2024}
}