> [!WARNING] > At some point we should check whether the same convexity properties hold for multiplicative perturbation. Here is a tweaked formulation compared to the InferOpt paper, which should make everything simpler. [perturbed_multiplicative.pdf](https://github.com/user-attachments/files/21167853/perturbed_multiplicative.pdf)
Warning
At some point we should check whether the same convexity properties hold for multiplicative perturbation.
Here is a tweaked formulation compared to the InferOpt paper, which should make everything simpler.
perturbed_multiplicative.pdf