🚀 DFL (Decision Focused Learning) is a new framework designed to enhance performance in downstream optimization tasks through prediction. Instead of simply reducing the prediction error, it focuses on minimizing the decision error. I am currently researching in this field, and I created this repository to share interesting and excellent papers on DFL for other researchers and practitioners.
- 🔥 I need Contributors!
| Date | keywords | Author | Paper | Publication |
|---|---|---|---|---|
| 2017 | PO | Priya L. Donti et al. | Task-based End-to-end Model Learning in Stochastic Optimization | NeurIPS Github |
| 2019 | CO | Bryan Wilder et al. | Melding the Data-Decisions Pipeline: Decision-Focused Learning for Combinatorial Optimization | AAAI Github |
| 2020 | PO | Jayanta Mandi and Tias Gubs | Interior Point Solving for LP-based prediction+optimisation | NeurIPS Github |
| 2021 | SPO | Adam N. Elmachtoub and Paul Grigas | Smart "Predict, then Optimize" | Management Science Github |
| 2022 | LTR | Jayanta Mandi et al. | Decision-Focused Learning: Through the Lens of Learning to Rank | ICML Spotlight Github |
| 2022 | PyEPO | Bo Tang and Elias B. Khalil | PyEPO: A Pytorch-based End-to-End Predict-then-Optimize Library for Linear and Interger Programming | arXiv Github |
| 2022 | LODL | Sanket Shah et al. | Decision-Focused Learning without Decision-Making: Learning Locally Optimized Decision Losses | NeurIPS Github |