YooJung Choi

Targeted Removal of Unwanted Biases in Probabilistic Circuits

A framework for bias removal in ML models helps institutions protect sensitive data and leverage domain-specific constraints in their use.

Program: MORE

Computing the Optimal Transport Map Between Probabilistic Circuits

Tractable computation of statistical distance metrics allows for theoretical guarantees on the robustness of machine learning models.

Program: FURI

Mitigating Label Bias Through Probabilistic Modeling With Latent Variable

Mitigating bias in datasets and certifying fairness in decisions using probabilistic circuits will help make machine learning tasks fairer.

Program: MORE