Jonas Arruda

The paper uses a Cellular Potts model and neural posterior estimation to infer cell migration dynamics in microfabricated environments, showing improved parameter inference over classical methods.

This study introduces an attack that reconstructs privacy-protected data in FL frameworks by exploiting covariance information and proposes defense strategies using differential privacy.

The paper introduces a machine learning approach to non-linear mixed-effects modeling using neural density estimation, demonstrating its flexibility and scalability in cell biology and pharmacology.