Abstract | ||
---|---|---|
We propose new multi-input shape constraints across four intuitive categories: complements, diminishers, dominance, and unimodality constraints. We show these shape constraints can be checked and even enforced when training machine-learned models for linear models, generalized additive models, and the nonlinear function class of multi-layer lattice models. Real-world experiments illustrate how the different shape constraints can be used to increase explainability and improve regularization, especially for non-IID train-test distribution shift. |
Year | Venue | DocType |
---|---|---|
2020 | ICML | Conference |
Citations | PageRank | References |
0 | 0.34 | 0 |
Authors | ||
6 |
Name | Order | Citations | PageRank |
---|---|---|---|
Maya R. Gupta | 1 | 595 | 49.62 |
Erez Louidor | 2 | 2 | 1.84 |
Oleksandr Mangylov | 3 | 0 | 0.34 |
Nobu Morioka | 4 | 0 | 0.34 |
Tamann Narayan | 5 | 0 | 0.34 |
Sen Zhao | 6 | 45 | 3.58 |