NOTE: Please check out ronaldseoh/bayesian-dl-experiments for my current Bayesian deep learning experiments!
This repository contains a fork of the original source codes for the experiments in the paper by Gal and Ghahramani "Dropout as a Bayesian Approximation: Representing Model Uncertainty in Deep Learning" (2015). This forked repository contains the original code plus experiment.ipynb - an adapted Jupyter Notebook version of experiment.py, and my own test results. Please visit here to see the original repo.
DropoutUncertaintyExps is licensed under Creative Commons Attribution-NonCommercial 4.0 International License. Please check LICENSE.