This repository contains a convolutional implementation of the described in Auto-Encoding Variational Bayes. The implemented model uses the MNIST dataset for classification in addition to the ADAM optimizer, batch normalization, weight decay, and ReLU non-linearities.
example.ipynb was written for a blog post
and shows a supervised and semi-supervised approach (using the VAE framework)
to classifying patients with benign or malignant tumors
Breast Cancer Wisconsin Diagnostic Data Set.
- Python 3.5 or greater
- Tensorflow 0.12.0 or greater