Skip to content

cchao0116/GSIMC-ICLR2023

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

4 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

GSIMC

License License Stars

The official implementation for "Graph Signal Sampling for Inductive One-Bit Matrix Completion: a Closed-form Solution".

What's news

[2023.04.26] We release the PySpark version of our codes for collaborative filtering.

Results on Netflix

Dataset

We use the Netflix benchmark to evaluate model performance, where the Dataframe scheme is as follows:

Feature Name Content
uid user identity
mid movie identity
time feedback time

CSV Download: Google, 夸克

Results with Rank=50

Below we report the HR@50, HR@100 and NDCG@100 results on the above provided dataset.

Graph Regularization Lapalacian HR@50 HR@100 NDCG@100
Bandlimited Norm Hypergraph 0.19623 0.29322 0.08761
Diffusion Process Hypergraph 0.18990 0.28547 0.08682
Random Walk Hypergraph 0.19512 0.29250 0.08766
Inverse Cosine Hypergraph 0.19353 0.29130 0.08757
Bandlimited Norm Covariance 0.20030 0.29517 0.08814
Diffusion Process Covariance 0.19228 0.28798 0.08675
Random Walk Covariance 0.19922 0.29500 0.08817
Inverse Cosine Covariance 0.19830 0.29540 0.08831

Folder Specification

  • conf/: configurations for logging
  • src/: codes for model definition
  • runme.sh: train or evaluate EasyDGL and baseline models

Run the Code

Download our data to $DATA_HOME directory, then Reproduce above results on Netflix benchmark:

bash runme.sh $DATA_HOME

Citation

If you find our codes useful, please consider citing our work

@inproceedings{chen2023graph,
  title={Graph Signal Sampling for Inductive One-Bit Matrix Completion: a Closed-form Solution},
  author={Chen, Chao and Geng, Haoyu and Zeng, Gang and Han, Zhaobing and Chai, Hua and Yang, Xiaokang and Yan, Junchi},
  booktitle={Proceedings of the International Conference on Learning Representations (ICLR'23)},
  year={2023},
}

About

Code for ICLR2023 paper "Graph Signal Sampling for Inductive 1-bit Matrix Completion: a Closed-Form Solution"

Topics

Resources

License

Stars

Watchers

Forks

Releases

No releases published