Graph Convolution RPCA with Adaptive Neighbors

This repository is the official implementation of GRPCA.

Requirements

To install requirements:

pip install -r requirements.txt

Preprocess the data

To better evaluate the performance of various methods, all the datasets are processed by following step:

1. Normalize the data.
2. Pollute the data. Sample 20% data points randomly and 20% features of them are reset by random values.

The raw data can be found in ./data, the processed data are in ./data_occ

Training and evaluation

To train and evaluate the model(s) in the paper, set the parameters in run.py, and run this command:

python run.py