Bayesean dot com yummy science

Intro Bayesian Inference With Stan

Introduction to Bayesian Inference (with Stan!)

I gave this talk at PyData NYC 2018.

Description

In this talk I’ll provide a quick, intuitive foundation for Bayesian inference and discuss how to use Stan to incorporate our knowledge into a statistical model. We will then perform inference on our data and see the limits of our knowledge reified.

Abstract

You’ve noticed that there are limits to data science. You’ve been surrounded by ad-hoc heuristics on nuanced datasets, and you’ve suspected for some time that this often leads to fragile inferences and poorly informed decision-making. You want more - you’d love to be able to incorporate more domain knowledge and really get a handle on what your data can actually tell you. Bayesian inference is the natural extension of logic to our uncertain, ever-changing world; rather than machine learning, it is machine-assisted human learning. In this talk I’ll provide a quick, intuitive foundation for Bayesian inference and discuss how to use Stan to incorporate our knowledge into a statistical model. We will then perform inference on our data and see the limits of our knowledge reified.

Slides