Philip-Bachman/Pseudo-Ensembles
Materials for implementing and reproducing results in the NIPS paper.
This code is for training feedforward neural networks with 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"
The useful and used parts of NN-Dropout
code for performing pragmatic banditry