A Score-Based Approach for Training Schrödinger Bridges for Data Modelling
Open Access
- 8 February 2023
- Vol. 25 (2), 316
- https://doi.org/10.3390/e25020316
Abstract
A Schrödinger bridge is a stochastic process connecting two given probability distributions over time. It has been recently applied as an approach for generative data modelling. The computational training of such bridges requires the repeated estimation of the drift function for a time-reversed stochastic process using samples generated by the corresponding forward process. We introduce a modified score- function-based method for computing such reverse drifts, which can be efficiently implemented by a feed-forward neural network. We applied our approach to artificial datasets with increasing complexity. Finally, we evaluated its performance on genetic data, where Schrödinger bridges can be used to model the time evolution of single-cell RNA measurements.This publication has 21 references indexed in Scilit:
- Delving Deep into Rectifiers: Surpassing Human-Level Performance on ImageNet ClassificationPublished by Institute of Electrical and Electronics Engineers (IEEE) ,2015
- On the Relation Between Optimal Transport and Schrödinger Bridges: A Stochastic Control ViewpointJournal of Optimization Theory and Applications, 2015
- A survey of the Schrödinger problem and some of its connections with optimal transportDiscrete & Continuous Dynamical Systems, 2014
- A Connection Between Score Matching and Denoising AutoencodersNeural Computation, 2011
- Exact and Computationally Efficient Likelihood-Based Estimation for Discretely Observed Diffusion Processes (with Discussion)Journal of the Royal Statistical Society Series B: Statistical Methodology, 2006
- Convergence of the Iterative Proportional Fitting ProcedureThe Annals of Statistics, 1995
- Reverse-time diffusion equation modelsStochastic Processes and their Applications, 1982
- Correlation functions and computer simulationsNuclear Physics B, 1981
- Derivation of the Schrödinger Equation from Newtonian MechanicsPhysical Review B, 1966
- A Relationship Between Arbitrary Positive Matrices and Doubly Stochastic MatricesThe Annals of Mathematical Statistics, 1964