For our Monte Carlo class , we reproduce the simulation experiments of the paper "On Multilevel Monte Carlo Unbiased Gradient Estimation for Deep Latent Variable Models" written by Yuyang Shi and Rob Cornish in 2021. We limit ourselves to the Linear Gaussian framework to quantify the bias and the variance of the computed estimators.
What you'll find in this repo:
- Final_Project.ipynb: our jupyter notebook reproducing the Linear Gaussian experiment (Yuyang Shi and Rob Cornish, 2021)