Perturbed Jacobians scale with the number of samples

JuliaDecisionFocusedLearning / InferOpt.jl

Combinatorial optimization layers for machine learning pipelines

https://juliadecisionfocusedlearning.github.io/InferOpt.jl/

MIT License

114 stars 4 forks source link

Perturbed Jacobians scale with the number of samples #74

Closed gdalle closed 1 year ago

gdalle commented 1 year ago

Reverse rules for PerturbedAdditive and PerturbedMultiplicative take a sum instead of a weighted mean. As a result we've been computing M * J instead of J this whole time