Philip-Bachman/Pseudo-Ensembles
Materials for implementing and reproducing results in the NIPS paper.
The useful and used parts of NN-Dropout
Public repository record indexed from GitHub. Explore verified star velocity metrics, source code repositories, and curated developer tool directories across the GitHubRepo ecosystem.
Materials for implementing and reproducing results in the NIPS paper.
Code for ICML 2014 paper: "Sample-based Approximate Regularization"
This code is for training feedforward neural networks with dropout.
code for performing pragmatic banditry