Reality gap

The difference between the real world and the simulations we use to model it

February 21, 2024 — November 2, 2024

approximation
Bayes
feature construction
likelihood free
machine learning
measure
metrics
probability
sciml
statistics
time series
Figure 1

Somewhere between approximating simulators and calibrating simulators, inference using simulators, and parameter search, we might wonder about the gap between them and the reality they claim to model. This gap might be an object of interest in itself.

This is a page to think about training models which characterise both the simulator and the real world, and to think about how to measure the gap between them.

1 Incoming

2 References

Brehmer, Louppe, Pavez, et al. 2020. “Mining Gold from Implicit Models to Improve Likelihood-Free Inference.” Proceedings of the National Academy of Sciences.
Higdon, Gattiker, Williams, et al. 2008. “Computer Model Calibration Using High-Dimensional Output.” Journal of the American Statistical Association.
Kennedy, and O’Hagan. 2001. “Bayesian Calibration of Computer Models.” Journal of the Royal Statistical Society: Series B (Statistical Methodology).
Koermer, Loda, Noble, et al. 2023. “Active Learning for Simulator Calibration.”
Mora, Yousefpour, Hosseinmardi, et al. 2024. “Neural Networks with Kernel-Weighted Corrective Residuals for Solving Partial Differential Equations.” arXiv Preprint arXiv:2401.03492.
Plumlee. 2017. “Bayesian Calibration of Inexact Computer Models.” Journal of the American Statistical Association.
Salvato, Fenu, Medvet, et al. 2021. “Crossing the Reality Gap: A Survey on Sim-to-Real Transferability of Robot Controllers in Reinforcement Learning.” IEEE Access.
Snoek, Larochelle, and Adams. 2012. “Practical Bayesian Optimization of Machine Learning Algorithms.” In Advances in Neural Information Processing Systems.
Tremblay, Prakash, Acuna, et al. 2018. “Training Deep Networks with Synthetic Data: Bridging the Reality Gap by Domain Randomization.” In 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition Workshops (CVPRW).