Kenji Fukumizu

DoubleGen: Debiased Generative Modeling of Counterfactuals

Conditionally Whitened Generative Models for Probabilistic Time Series Forecasting

Generalized Sobolev IPM for Graph-Based Measures

An Efficient Orlicz-Sobolev Approach for Transporting Unbalanced Measures on a Graph

Pairwise Optimal Transports for Training All-to-All Flow-Based Condition Transfer Model

Compositional simulation-based inference for time series

Flow matching achieves almost minimax optimal convergence

Scalable Sobolev IPM for Probability Measures on a Graph

Extended Flow Matching : a Method of Conditional Generation with Generalized Continuity Equation