shehel / Bayesian_Learning

Probabilistic programming - Bayesian deep networks and GPs

• GP Regression: Introduction to Gaussian processes with a regression problem using PyMC3 PPL.
• MC Dropout: Uncertainty estimates usng dropout.
• Bayesian CNN: Implementing a Bayesian CNN for classification using variational inference and MNIST with Edward PPL. Includes uncertainty estimation on nMNIST.
• Bayesian Regression with Neural Networks: Non-linear regression with neural networks using Variational Inference. Compares working with VI in Edward and PyMC3 PPLs