When Osvaldo asked me to write the foreword to his new book I felt honored, excited, and a bit scared, so naturally I accepted. What follows is my best attempt to convey what makes probabilistic programming so exciting to me. Osvaldo did a great job with the book, it is …

# Using Bayesian Decision Making to Optimize Supply Chains

(c) 2019 Thomas Wiecki & Ravin Kumar

As advocates of Bayesian statistics in data science we often have to convince business-minded colleagues or customers of the added value of such an approach. While there are many good reasons for applying Bayesian modeling to solve business problems (Sean J Taylor recently had …

# Hierarchical Bayesian Neural Networks with Informative Priors

(c) 2018 by Thomas Wiecki

Imagine you have a machine learning (ML) problem but only small data (*gasp*, yes, this does exist). This often happens when your data set is nested -- you might have many data points, but only few per category. For example, in ad-tech you may want predict …

# An intuitive, visual guide to copulas

(c) 2018 by Thomas Wiecki

People seemed to enjoy my intuitive and visual explanation of Markov chain Monte Carlo so I thought it would be fun to do another one, this time focused on copulas.

If you ask a statistician what a copula is they might say "a copula is …

# What's new in PyMC3 3.1

We recently released PyMC3 3.1 after the first stable 3.0 release in January 2017. You can update either via `pip install pymc3`

or via `conda install -c conda-forge pymc3`

.

A lot is happening in PyMC3-land. One thing I am particularily proud of is the developer community we have …

# Random-Walk Bayesian Deep Networks: Dealing with Non-Stationary Data

Download the NB: https://github.com/twiecki/WhileMyMCMCGentlySamples/blob/master/content/downloads/notebooks/random_walk_deep_net.ipynb

(c) 2017 by Thomas Wiecki -- Quantopian Inc.

Most problems solved by Deep Learning are stationary. A cat is always a cat. The rules of Go have remained stable for 2,500 years, and will likely …

# Why hierarchical models are awesome, tricky, and Bayesian

(c) 2017 by Thomas Wiecki

Hierarchical models are underappreciated. Hierarchies exist in many data sets and modeling them appropriately adds a boat load of statistical power (the common metric of statistical power). I provided an introduction to hierarchical models in a previous blog post: Best Of Both Worlds: Hierarchical Linear …

# Bayesian Deep Learning Part II: Bridging PyMC3 and Lasagne to build a Hierarchical Neural Network

(c) 2016 by Thomas Wiecki

Recently, I blogged about Bayesian Deep Learning with PyMC3 where I built a simple hand-coded Bayesian Neural Network and fit it on a toy data set. Today, we will build a more interesting model using Lasagne, a flexible `Theano`

library for constructing various types of …

# Bayesian Deep Learning

# Variational Inference: Bayesian Neural Networks¶

(c) 2016-2018 by Thomas Wiecki, updated by Maxim Kochurov

Original blog post: https://twiecki.github.io/blog/2016/06/01/bayesian-deep-learning/

## Current trends in Machine Learning¶

There are currently three big trends in machine learning: **Probabilistic Programming**, **Deep Learning** and "**Big Data**". Inside of PP …

# MCMC sampling for dummies

When I give talks about probabilistic programming and Bayesian statistics, I usually gloss over the details of how inference is actually performed, treating it as a black box essentially. The beauty of probabilistic programming is that you actually don't *have* to understand how the inference works in order to build …