Factorization Machine models in PyTorch

Accelerating Inference for Recommendation Systems (WSDM'21)

A ready-to-use framework of the state-of-the-art models for structured (tabular) data learning with PyTorch. Applications include recommendation, CRT prediction, healthcare analytics, anomaly detection, and etc.

Privacy-Preserving Bandits (MLSys'20)

Acceleration and Compression of DeepFwFM.

A decision-focused uplift modeling framework that jointly optimizes CATE prediction and treatment allocation policy via a shared-layer neural network, benchmarked against S-Learner, X-Learner, UpliftRank, and GRF on CRITEO-UPLIFT v2.