FMMI: Flow Matching Mutual Information Estimation
By: Ivan Butakov , Alexander Semenenko , Alexey Frolov and more
Potential Business Impact:
Makes computers understand data better and faster.
We introduce a novel Mutual Information (MI) estimator that fundamentally reframes the discriminative approach. Instead of training a classifier to discriminate between joint and marginal distributions, we learn a normalizing flow that transforms one into the other. This technique produces a computationally efficient and precise MI estimate that scales well to high dimensions and across a wide range of ground-truth MI values.
Similar Papers
A Neural Difference-of-Entropies Estimator for Mutual Information
Machine Learning (Stat)
Helps computers understand how things are connected.
MIST: Mutual Information Via Supervised Training
Machine Learning (CS)
Teaches computers to guess how much two things are related.
InfoBridge: Mutual Information estimation via Bridge Matching
Machine Learning (CS)
Finds hidden connections in data better.