A no non-sense guide to get you started quickly with Bayesian Modeling

Designed specifically for busy professionals seeking practical mastery of Bayesian inference and modeling fundamentals.
The curriculum strikes an optimal balance between theoretical concepts and real-world applications, with knowledge delivered through an engaging mix of practical examples, executable code, and hands-on exercises.
You'll gain proficiency in core Bayesian methodologies including inferential techniques, A/B testing frameworks, change point detection algorithms, conjugate prior relationships, and Markov Chain Monte Carlo simulation methods.

From basics to advanced concepts step-by-step

📚 Comprehensive foundation in Bayesian statistical principles and methodology

📈 Expert guidance on prior selection strategies and best practices

🧠 Rigorous approaches to hypothesis testing and ensemble modeling techniques

👍 In-depth exploration of essential Bayesian inferential frameworks

👓 Clear, intuitive explanations of Markov Chain Monte Carlo methods

⚙️ Hands-on implementation of hierarchical models from first principles

💻 Practical programming workshops for Metropolis and Gibbs sampling algorithms

⚡ Advanced techniques for developing efficient conjugate samplers

Why this course?

Crafted for practitioners seeking to fast-track their mastery of Bayesian methodologies. Dive into essential theoretical foundations, discover practical applications, and develop the technical skills to build custom models and implement sophisticated sampling algorithms.

  • 8 Hours of Content

    Professionally edited, designed. Developed based on learning best practices.

  • Github Starter Code

    Download the code locally or work from a "Binder" notebook directly online.

  • Well-Documented Examples

    Examples are clear and intuitive.

Getting Started

This section focuses on getting set up and the importance of Bayesian methods.

Getting Started

How to Succeed with this Course!

Where to Get the Code?

Being a Bayesian

Where to Begin - Course Outline

Enroll - let's get started!

Part 1: Bayesian Foundations

Here we explore how Bayesian learning occurs and review the key concepts underlying Bayesian inference and modeling

Lecture 1: Motivating Example

Lecture 2: Bayes Theorem Part 1

Lecture 3: Bayes Theorem Part 2

Lecture 4: Bayes Theorem Part 3

Lecture 5: Programming: Beta Prior with a Binomial Likelihood

Lecture 6: Exercise - Applying Bayes Theorem

Lecture 7: Programming: Gamma Prior and Poisson Likelihood

Lecture 8: Programming: Modularizing Code and Misspecified Priors

Lecture 9: Prior Distributions

Lecture 10: Programming: The effect of priors on Posteriors

Lecture 11: Posterior Distributions Part 1

Lecture 12: Programming: Transforms, MAPs and CIs

Lecture 13: Posterior Distributions Part 2

Lecture 14: Programming: Combining Priors and Posteriors

Lecture 15: Posterior Predictive Distributions Part 1

Lecture 16: Programming: Exploring the Posterior Predictive Distribution

Lecture 17: Posterior Predictive Distributions Part 2

Part 2: Conjugate Distributions

This section we develop the core tools to be able to build fast samplers by leveraging the conjugate property. In addition we review how Hierarchical modeling is related to regularization.

Lecture 18: Intro to Conjugate Distributions

Lecture 19: Conjugacy Explained + Exponential Families Part 1

Lecture 20: Conjugacy Explained + Exponential Families Part 2

Lecture 21: The "Normal" Model

Lecture 22: The Beta-Binomial Model

Lecture 23: Hierarchical Model Theory: Part 1

Lecture 24: Hierarchical Model Theory: Part 2

Lecture 25: Hierarchical Model Application: Part 1

Lecture 26: Hierarchical Model Application: Part 2

Lecture 27: Programming a Hierarchical Model

Lecture 28: Coming up next - MCMC Samplers

Part 3: MCMC Samplers

In this section we review the key components of Markov Chain Monte Carlo and how to construct samplers for our posterior distributions.

Lecture 29: Intro to Markov Chain Monte Carlo (MCMC)

Lecture 30: Review of MCMC Part 1

Lecture 31: Review of MCMC Part 2

Lecture 32: Review of MCMC Part 3

Lecture 33: MCMC Diagnostics Part 1

Lecture 34: MCMC Diagnostics Part 2

Lecture 35: Metropolis Sampler: How it Works

Lecture 36: Metropolis Sampler: Normal Model

Lecture 37: Metropolis Sampler: Poisson-Gamma Model

Lecture 38: Metropolis Sampler: Poisson-Gamma Model - Code

Lecture 39: Gibbs Sampler + Normal Model

Lecture 40: Gibbs Sampler: Why it Works

Lecture 41: Gibbs Sampler: A Switch Point Model

Lecture 42: Metropolis and Gibbs Example + Diagnostics Part 1

Lecture 43: Metropolis and Gibbs Example + Diagnostics Part 2

Lecture 44: Other Types of Samplers

Lecture 45: Where to from here?

What people are saying

Excellent course!

Basil Singer, PhD, FSA, CERA. CEO of Modellicity

Bayesian Foundation for Professionals gave me a lot of insight in the field of Bayesian statistics, which I’ve been lacking. I strongly recommend taking this course for anyone who wishes to enhance their understanding of mechanics of Bayesian statistics and how it can be applied in practice. Looking forward to the next one in the series!

Bayesian Foundations for Professionals provides an excellent overview and functional primer on using bayesian methods

Dylan Duchen, PhD, Yale-Boehringer Ingelheim Biomedical Data Science Fellow

With easy to follow explanatory videos and excellently annotated code/programming modules, I'm incredibly happy with the course and look forward to using these methods in my own work. Excited to view Dr. Touyz's future courses!

Strong course directed at an important group, working data scientists who haven't used Bayesian methods before

Steve Haptonstahl, PhD, Director, Transfix

Making code so prominent is a great choice for a practitioner -- we all know that there's no better way to be specific than to write and read actual code.

delivers on its promise by providing aspiring and seasoned data science professionals an opportunity to upgrade and sharpen their skills

Jarvis Joyce, Chief Computational Linguist, PeopleBeforeCode Inc.

Bayesian Foundations for Professionals - Theory and Applications delivers on its promise by providing aspiring and seasoned data science professionals an opportunity to upgrade and sharpen their skills. Josh’s professorial background shines as he provides clear instruction and challenges course takers to apply Bayesian knowledge to their respective applications. UpSkill ML’s first course is certainly worth taking!

FAQ

  • Who should take this course?

    This course is designed for technical professionals that are seeking to develop a foundational understanding of Bayesian Methods.

  • What coding language do you use throughout the course?

    This is a python based course however many of the ideas are portable to other programming languages with statistical libraries. We mainly use numpy, pandas, scipy and plotnine throughout the course so familiarity with these libraries is helpful however not required.

  • What are the pre-requisites for this course?

    A background in statistics, common probability distributions and python programming are helpful for navigating the topics in this course. While they aren’t hard pre-requisites to understanding the material they will certainly make it easier.

  • Do I get a certificate of completion?

    This is a feature that will be available in the future - at the moment we don’t support issuing certificates of completion.

  • What will I be able to do after this course?

    After this course you will have the tools to develop your expertise and explain the ideas at a foundational level. You will be able to speak informatively about Bayesian methods and design your own samplers through the concepts taught in this course.

  • Do you offer refunds?

    Refund requests will be reviewed for reimbursement. A valid request must be submitted within 7 days from the time of purchase. Eligible refunds will be calculated as the cost of the course minus administrative overhead for processing. Note: refunds cannot be issued to those who purchased the course at a discounted rate.

  • Who do I email with questions?

    For questions or more information please email [email protected].