Factorization Machine models in PyTorch

This package provides a PyTorch implementation of factorization machine models and common datasets in CTR prediction.

Available Datasets

• MovieLens Dataset
• Criteo Display Advertising Challenge
• Avazu Click-Through Rate Prediction

Available Models

Model	Reference
Logistic Regression
Factorization Machine	S Rendle, Factorization Machines, 2010.
Field-aware Factorization Machine	Y Juan, et al. Field-aware Factorization Machines for CTR Prediction, 2015.
Higher-Order Factorization Machines	M Blondel, et al. Higher-Order Factorization Machines, 2016.
Factorization-Supported Neural Network	W Zhang, et al. Deep Learning over Multi-field Categorical Data - A Case Study on User Response Prediction, 2016.
Wide&Deep	HT Cheng, et al. Wide & Deep Learning for Recommender Systems, 2016.
Attentional Factorization Machine	J Xiao, et al. Attentional Factorization Machines: Learning the Weight of Feature Interactions via Attention Networks, 2017.
Neural Factorization Machine	X He and TS Chua, Neural Factorization Machines for Sparse Predictive Analytics, 2017.
Neural Collaborative Filtering	X He, et al. Neural Collaborative Filtering, 2017.
Field-aware Neural Factorization Machine	L Zhang, et al. Field-aware Neural Factorization Machine for Click-Through Rate Prediction, 2019.
Product Neural Network	Y Qu, et al. Product-based Neural Networks for User Response Prediction, 2016.
Deep Cross Network	R Wang, et al. Deep & Cross Network for Ad Click Predictions, 2017.
DeepFM	H Guo, et al. DeepFM: A Factorization-Machine based Neural Network for CTR Prediction, 2017.
xDeepFM	J Lian, et al. xDeepFM: Combining Explicit and Implicit Feature Interactions for Recommender Systems, 2018.
AutoInt (Automatic Feature Interaction Model)	W Song, et al. AutoInt: Automatic Feature Interaction Learning via Self-Attentive Neural Networks, 2018.
AFN(AdaptiveFactorizationNetwork Model)	Cheng W, et al. Adaptive Factorization Network: Learning Adaptive-Order Feature Interactions, AAAI'20.

Each model's AUC values are about 0.80 for criteo dataset, and about 0.78 for avazu dataset. (please see example code)

Installation

pip install torchfm

API Documentation

https://rixwew.github.io/pytorch-fm

Licence

MIT