diff --git a/task1/readme.md b/task1/readme.md index bd52f80..ff70971 100644 --- a/task1/readme.md +++ b/task1/readme.md @@ -46,7 +46,7 @@ Estimate quality of approximation, speed of approximation, and KL divergence bet * [hyperparameter optimization](http://strijov.com/papers/HyperOptimizationEng.pdf) * Visualization: [help1](https://matplotlib.org/stable/users/interactive.html),[help2](https://stackoverflow.com/questions/44329068/jupyter-notebook-interactive-plot-with-widgets),[help3](https://towardsdatascience.com/matplotlib-animations-in-jupyter-notebook-4422e4f0e389) -8. [Reporter: enter your name] Compare approximation with Laplace approximation, Variational inference and Expectation propagation for: +8. [Reporter: Timofey Chernikov] Compare approximation with Laplace approximation, Variational inference and Expectation propagation for: - Gaussian mixture (visualize for different parameters of mixture) - Beta-distribution with a and beta < 1 (visualize for different parameters) * [Bishop: describes all the methods](https://www.microsoft.com/en-us/research/uploads/prod/2006/01/Bishop-Pattern-Recognition-and-Machine-Learning-2006.pdf)