Towards Interpretable Multi-Task Learning Using Bilevel Programming

Abstract

Interpretable Multi-Task Learning can be expressed as learning a sparse graph of the task relationship based on the prediction performance of the learned models. Since many natural phenomenon exhibit sparse structures, enforcing sparsity on learned models reveals the underlying task relationship. Moreover, different sparsification degrees from a fully connected graph uncover various types of structures, like cliques, trees, lines, clusters or fully disconnected graphs. In this paper, we propose a bilevel formulation of multi-task learning that induces sparse graphs, thus, revealing the underlying task relationships, and an efficient method for its computation. We show empirically how the induced sparse graph improves the interpretability of the learned models and their relationship on synthetic and real data, without sacrificing generalization performance. Code at this https URL

Cite

Text

Alesiani et al. "Towards Interpretable Multi-Task Learning Using Bilevel Programming." European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, 2020. doi:10.1007/978-3-030-67661-2_35

Markdown

[Alesiani et al. "Towards Interpretable Multi-Task Learning Using Bilevel Programming." European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, 2020.](https://mlanthology.org/ecmlpkdd/2020/alesiani2020ecmlpkdd-interpretable/) doi:10.1007/978-3-030-67661-2_35

BibTeX

@inproceedings{alesiani2020ecmlpkdd-interpretable,
  title     = {{Towards Interpretable Multi-Task Learning Using Bilevel Programming}},
  author    = {Alesiani, Francesco and Yu, Shujian and Shaker, Ammar and Yin, Wenzhe},
  booktitle = {European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases},
  year      = {2020},
  pages     = {593-608},
  doi       = {10.1007/978-3-030-67661-2_35},
  url       = {https://mlanthology.org/ecmlpkdd/2020/alesiani2020ecmlpkdd-interpretable/}
}