Login
Intuitive Bayes/Advanced Regression
Buy now

Advanced Regression

  • Course
  • 143 Lessons

Bayesian regression can solve a variety of real world problems. In this course, we'll use modern packages such as PyMC and Bambi. We'll answer questions about product pricing, election estimation, and more.
Sign up Here

Syllabus

About this Course

Why we created it, who's it for, and what to expect
Welcome to Advanced Regression
Preview
The Problem with Ordinary Regression
Preview
Bayesian Generalized Linear Models
Preview
What's Different Now
Preview
Who This Course is For
Preview
Outline and Prequisites
Preview
Course Presurvey
Lesson Feedback
Meet your Instructors

Resource Library

Instructions on how access the various resources provided in this course
Intuitive Bayes Discourse Community
Github Repository and Code Access
Environment Installation with Anaconda
Preview
Optional Orientation: Discourse
Preview
Optional Orientation: Github
Preview
Optional Orientation: Podia
Preview

Regression Refresher

Revisiting the fundamentals
Introduction
Preview
Exploratory Data Analysis
Preview
Making a Plan
Preview
The World's Simplest Model
Preview
Adding a Slope
Preview
Transformations
Preview
Accounting for Species
Preview
When We Catch New Fish
Preview
From Predictions to Insights
Preview
Bayesian Workflow and Growing Pains
Preview
Lesson Recap
Preview
Lesson Feedback
Lesson References
Lesson Exercises

Bambi Deep Dive

In this lesson we'll introduce you to Bambi, an incredibly useful tool for Bayesian practitioners
Introduction
The World's Simplest Model, Now Simpler
Peeking Under The Hood
Sloping Up
Transformations in Bambi
Modeling Categories
Parameter Identifiability
Understanding Encodings
The Full Model
Predictions
An End to End Trip with Bambi
Lesson Recap
Lesson Feedback
Lesson Exercises
Lesson References

Advanced AB Testing and Link Functions

Taking regression to the next level
Introduction
Basic AB Test
AB Testing with Continuous Covariates
Link Functions and Generalized Linear Models
AB Testing End to End
United States Election Forecasting
Lesson Recap
Lesson Feedback Form
Lesson Exercises
Lesson References

Categorical Regression

When there's more than 2 options and you don't know what to do!

Introduction
Nerdy crêpes
Our First Categorical model
Killing Me Softmax
Running our Categorical regression
Adding Regressors
Fun with Shapes
Unpleasant Probabilities
Sampling & Diagnosing
Debugging our model
Categorical model with Bambi
Introducing pm.ZeroSumNormal
SALK EDA
Coding the Salk model in PyMC
Sampling the Salk model
Estonia, Bambi styyyyyle!
Lesson Recap
Lesson Exercises
Lesson References
Lesson Feedback

Multinominal Regression

How aggregating counts allows for even more powerful estimations

Introduction
Aggregating the Estonian data
Introducing geographical variables
Writing the Multinomial model
Coding the Model
Sampling & Posterior analysis
Taking all effects into account
Posterior Schematics
Taking sample size into account
Defining & Sampling the Bambi Multinomial model
Analyzing the posterior effects
Computing the latent probabilities
Comparing PyMC and Bambi estimates
Posterior Retrodictive Sampling
Lesson Recap
Lesson Exercises
Lesson References
Lesson Feedback

Counting Things

Going back to the basics

Introduction
Counting Things
Counting Regression
Poisson Distributions in Sports
Soccer Statistics
Home Field Advantage
Zero Inflated Poisson
Zero Inflated Fish
Lesson Recap
Lesson Exercises
Lesson Feedback
Lesson References

Overdispersed Regression

When distributions need some extra help

12 segundos de oscuridad
Unanticipated consequences
What's the deal with overdispersion
When the whole is more than the sum of the parts
Negative binomial regression
Overdispersion for all
Beta-binomial to the rescue
The beta-binomial model in action
Counting grasshoppers
Beta-binomial for grasshoppers
Lesson Recap
Lesson Feedback
Lesson References
Lesson Exercises

Hierarchical Models

Introduction
Pulling and Unpooling Extremities
Simple Hierarchical Model
Regularization to the mean
Complete Hierarchical Model
Non-centered hierarchical model
Sampling & Posterior Analysis
In-sample Predictions
Analysis of Variance
LOO comparison
Predicting on new groups
Estonian Hierarchy
Writing down the Hierarchical Model
Coding the Full Model
Hierarchical Multinomial, Deluxe Version
Sampling, Posterior & Post-stratification
Post-stratification in practice
Making post-stratified predictions
Visualizing Strata Predictions
A Deeper Cut
Lesson Recap
Lesson Feedback
Lesson References
Lesson Exercises

Final Notes

Congratulations