This code is the official implementation of the paper: MCM: Masked Cell Modeling for Anomaly Detection in Tabular Data published at ICLR 2024 as a conference paper by Jiaxin Yin, Yuanyuan Qiao, Zitang Zhou, Xiangchao Wang, and Jie Yang. The code allows the users to reproduce and extend the results reported in the study. Please cite the above paper when reporting, reproducing or extending the results.
- When using your own data, move the dataset into
./Data. - Add the dataset name to
./Dataset/DataLoader.pybased on the format of your dataset. - Modify dataset_name and data_dim in
./main.py - You can download tabular datasets from ODDS and ADBench for testing.
Run main.py to start training and testing the model. Results will be automatically stored in ./results.
- Python 3.6
- PyTorch 1.10.2
- torchvision 0.11.3
- numpy 1.23.5
- pandas 1.5.3
- scipy 1.10.1