The PyTorch implementation of paper REST: Debiased Social Recommendation via Reconstructing Exposure Strategies

Last update: Mar 13, 2022

Related tags

Deep Learning REST

Overview

REST

The PyTorch implementation of paper REST: Debiased Social Recommendation via Reconstructing Exposure Strategies.

Usage

Download dataset

Download datasets: Ciao, Epinions, Yelp

Then unzip them into the directory datasets and run preprocess_yelp.py.

└── datasets
    ├── Ciao
    │   ├── rating_with_timestamp.mat
    │   ├── trust.mat
    ├── Epinions
    │   ├── rating_with_timestamp.mat
    │   ├── trust.mat
    ├── Yelp
    │   ├── yelp_academic_dataset_review.json
    │   ├── yelp_academic_dataset_user.json
    │   ├── noiso_reid_u2uir.npz
    │   ├── ...

Run

    python ./run_rate/run_rest_rate_ciao.py

During the training, we can obtain some logs and model-checkpoints in the directory logs and saved_models,

Results

Model	Ciao RMSE	Epinions RMSE	Yelp RMSE
PMF	1.1936±0.0019	1.2755±0.0022	1.2454±0.0011
NeuMF	0.9828±0.0022	1.0838±0.0015	1.1958±0.0005
MultiVAE	1.1908±0.0014	1.2104±0.0039	1.2944±0.0020
RecVAE	1.1787±0.0022	1.1946±0.0038	1.2385±0.0014
CausE	1.0003±0.0013	1.0705±0.0013	1.2039±0.0015
CVIB-MF	1.2001±0.0011	1.2477±0.0003	1.3189±0.0024
CVIB-NCF	1.0462±0.0013	1.2477±0.0003	1.3613±0.0043
MACR-MF	1.1859±0.0030	1.2364±0.0031	1.2344±0.0004
DecRS	0.9875±0.0033	1.0617±0.0033	-
GraphRec	0.9743±0.0021	1.0567±0.0019	1.1968±0.0017
NGCF	1.0135±0.0010	1.1286±0.0017	1.2231±0.0017
LightGCN	1.1919±0.0014	1.2025±0.0005	1.2444±0.0019
REST	0.9635±0.0009	1.0413±0.0007	1.1733±0.0006

Detailed results can be found in the paper.

The PyTorch implementation of paper REST: Debiased Social Recommendation via Reconstructing Exposure Strategies

Related tags

Overview

REST

Usage

Download dataset

Run

Results

Owner

DMIRLAB

Algorithmic encoding of protected characteristics and its implications on disparities across subgroups

Hierarchical Memory Matching Network for Video Object Segmentation (ICCV 2021)

Automatic Calibration for Non-repetitive Scanning Solid-State LiDAR and Camera Systems

DIRL: Domain-Invariant Representation Learning

Code for CoMatch: Semi-supervised Learning with Contrastive Graph Regularization

SeqFormer: a Frustratingly Simple Model for Video Instance Segmentation

HyperPose is a library for building high-performance custom pose estimation applications.

Convert weight file.pth to weight file.blob

EssentialMC2 Video Understanding

4D Human Body Capture from Egocentric Video via 3D Scene Grounding

Pytorch implementation of our paper accepted by NeurIPS 2021 -- Revisiting Discriminator in GAN Compression: A Generator-discriminator Cooperative Compression Scheme

Rafael Project- Classifying rockets to different types using data science algorithms.

Task-related Saliency Network For Few-shot learning

PyTorch implementation of U-TAE and PaPs for satellite image time series panoptic segmentation.

통일된 DataScience 폴더 구조 제공 및 가상환경 작업의 부담감 해소

[ICRA2021] Reconstructing Interactive 3D Scene by Panoptic Mapping and CAD Model Alignment

Adversarial Reweighting for Partial Domain Adaptation

Causal Imitative Model for Autonomous Driving

A simple tutoral for error correction task, based on Pytorch

Reference implementation for Deep Unsupervised Learning using Nonequilibrium Thermodynamics