Large Scale Tensor Regression Using Kernels and Variational Inference
Abstract
We outline an inherent flaw of tensor factorization models when latent factors are expressed as a function of side information and propose a novel method to mitigate this. We coin our methodology kernel fried tensor (KFT) and present it as a large-scale prediction and forecasting tool for high dimensional data. Our results show superior performance against LightGBM and Field aware factorization machines (FFM), two algorithms with proven track records, widely used in large-scale prediction. We also develop a variational inference framework for KFT which enables associating the predictions and forecasts with calibrated uncertainty estimates on several datasets.
Cite
Text
Hu et al. "Large Scale Tensor Regression Using Kernels and Variational Inference." Machine Learning, 2022. doi:10.1007/S10994-021-06067-7Markdown
[Hu et al. "Large Scale Tensor Regression Using Kernels and Variational Inference." Machine Learning, 2022.](https://mlanthology.org/mlj/2022/hu2022mlj-large/) doi:10.1007/S10994-021-06067-7BibTeX
@article{hu2022mlj-large,
title = {{Large Scale Tensor Regression Using Kernels and Variational Inference}},
author = {Hu, Robert and Nicholls, Geoff K. and Sejdinovic, Dino},
journal = {Machine Learning},
year = {2022},
pages = {2663-2713},
doi = {10.1007/S10994-021-06067-7},
volume = {111},
url = {https://mlanthology.org/mlj/2022/hu2022mlj-large/}
}