Robustness in Sum-Product Networks with Continuous and Categorical Data
Abstract
Sum-product networks are a popular family of probabilistic graphical models for which marginal inference can be performed in polynomial time. After learning sum-product networks from scarce data, small variations of parameters could lead to different conclusions. We adapt the robustness measure created for categorical credal sum-product networks to domains with both continuous and categorical variables. We apply this approach to a real-world dataset of online purchases where the goal is to identify fraudulent cases. We empirically show that such credal models can better discriminate between easy and hard instances than simply using the probability of the most probable class.
Cite
Text
de Wit et al. "Robustness in Sum-Product Networks with Continuous and Categorical Data." Proceedings of the Eleventh International Symposium on Imprecise Probabilities: Theories and Applications, 2019.Markdown
[de Wit et al. "Robustness in Sum-Product Networks with Continuous and Categorical Data." Proceedings of the Eleventh International Symposium on Imprecise Probabilities: Theories and Applications, 2019.](https://mlanthology.org/isipta/2019/dewit2019isipta-robustness/)BibTeX
@inproceedings{dewit2019isipta-robustness,
title = {{Robustness in Sum-Product Networks with Continuous and Categorical Data}},
author = {de Wit, Rob and de Campos, Cassio P. and Conaty, Diarmaid and del Rincon, Jesus Martinez},
booktitle = {Proceedings of the Eleventh International Symposium on Imprecise Probabilities: Theories and Applications},
year = {2019},
pages = {156-158},
volume = {103},
url = {https://mlanthology.org/isipta/2019/dewit2019isipta-robustness/}
}