Streaming-Bayesian-Deep-Tensor

by Shikai Fang, Zheng Wang, Zhimeng Pan, Ji Liu, Shandian Zhe

code for paper : Streaming Bayesian Deep Tensor Factorization for ICML 2021

Links for Paper and Supplementary

Requirements:

This project is implemented based on the framework of Probabilistic-Backpropagation ( Paper), which applies the Theano and python 2.7. Please find the requirement details in PBP project, make sure you insatll Theano and can run the PBP bug-free.

Instructions:

1. Clone this repository.
2. See model details in SBDT_net\ . Run model with $ python SBDT_datasetname.py

Datasets & Baselines

We offer three large size tensor dataset for test.
Binary-data dataset: Anime at .\stream_nn_td_binary\data_binary\anime
Real-data dataset: Acc at .\stream_ss_td_real\data_real\acc and
Movilens1M at .\stream_ss_td_real\data_real\movielens_1m

Check our paper for more datasets and baselines.

Citation

Please cite our work if you would like it

@inproceedings{fang2021streaming,
  title={Streaming Bayesian Deep Tensor Factorization},
  author={Fang, Shikai and Wang, Zheng and Pan, Zhimeng and Liu, Ji and Zhe, Shandian},
  booktitle={International Conference on Machine Learning},
  pages={3133--3142},
  year={2021},
  organization={PMLR}
}