Probabilistic Machine Learning Series Post 3

Probabilistic Machine Learning Series Post 3: Weights Uncertainty with Correlated Noise

Using correlated dropout to quantify the uncertainty of a Neural Network forecasts

One of the main drawbacks of the Bayesian approach is that it is not scalable. In 2015, Blundell & al. proposed to quantify the uncertainty in Neural Networks by using dropout. The method, called Bayes by backprop, uses the reparametrization trick to approximate a variational bound to the posterior. In this post, we will experiment …

bidirectional-lstm data science hands-on-tutorials machine machine-learing machine learning probabilistic-models series