From PBDN

Introduction

The general idea of variance reduction as applied in this project is to introduce dependencies between trajectories, but not within an individual trajectory. Each individual trajectory behaves according to the rules of the model and the simulation algorithm, but together, fewer of them are needed to get to the final solution than would be the case if no variance reduction was used.

The four variance reduction techniques examined were:

• Antithetic variates
• Stratified sampling
• Importance sampling
• Randomized quasi-Monte Carlo

Back to Simulation of Biochemical Reaction Networks main page.