Projection losses

Python implementation of "Structured Prediction with Projection Oracles".

Supported polytopes

• Probability simplex
• Unit cube
• Knapsack polytope
• Birkhoff polytope
• Permutahedron
• Order simplex
• Cartesian products

Installation

Simply copy relevant files to your project.

References

[1]	SparseMAP: Differentiable Sparse Structured Inference. Vlad Niculae, André F. T. Martins, Mathieu Blondel, Claire Cardie. In Proc. of ICML 2018. [arXiv]

[2]	Learning Classifiers with Fenchel-Young Losses: Generalized Entropies, Margins, and Algorithms. Mathieu Blondel, André F. T. Martins, Vlad Niculae. In Proc. of AISTATS 2019. [arXiv]

[3]	Learning with Fenchel-Young Losses. Mathieu Blondel, André F. T. Martins, Vlad Niculae. Preprint. [arXiv]

[4]	Structured Prediction with Projection Oracles. Mathieu Blondel. In Proc. of NeurIPS 2019. [arXiv]

Author

• Mathieu Blondel, 2019