Monthly Subscription Starting at AED 99 + VAT

Bayesian Inference with MCMC
  • Level Foundation
  • Duration 15 hours
  • Course by Databricks
  • Offered by

About

The objective of this course is to introduce Markov Chain Monte Carlo Methods for Bayesian modeling and inference, The attendees will start off by learning the the basics of Monte Carlo methods. This will be augmented by hands-on examples in Python that will be used to illustrate how these algorithms work. This will be the second course in a specialization of three courses .Python and Jupyter notebooks will be used throughout this course to illustrate and perform Bayesian modeling with PyMC3. The course website is located at https://sjster.github.io/introduction_to_computational_statistics/docs/index.html. The course notebooks can be downloaded from this website by following the instructions on page https://sjster.github.io/introduction_to_computational_statistics/docs/getting_started.html. The instructor for this course will be Dr. Srijith Rajamohan.

Modules

Metrics to assess model performance - I

    • 8 Videos
    • 2 Readings
Show info about module content

8 Videos

  • Welcome to Course 2!
  • Introduction
  • Underfitting and Overfitting
  • Explained Variance
  • Cross Validation
  • Information Criteria
  • Log-likelihood and Deviance
  • Posterior Predictive Distribution

2 Readings

  • What can you expect from this course/specialization?
  • Likelihood and its use in Parameter Estimation and Model Comparison

Metrics to assess model performance - II

    • 5 Videos
    • 3 Readings
    • 1 Assignment
Show info about module content

1 Assignment

  • Topics in Model Performance

5 Videos

  • AIC, BIC, DIC and WAIC
  • A qualitative discussion of the various metrics
  • Entropy
  • KL Divergence
  • Model Averaging

3 Readings

  • Understanding predictive information criteria for Bayesian models
  • Information Theory and Statistics
  • Model Stacking

The Foundations of Bayesian Inference

    • 3 Videos
    • 1 Readings
Show info about module content

3 Videos

  • Introduction
  • Markov Chains
  • Why does Markov Chain Monte Carlo work?

1 Readings

  • Markov Chains

The Metropolis Algorithm

    • 4 Videos
    • 1 Plugin
Show info about module content

4 Videos

  • The Metropolis algorithm for sampling
  • The Metropolis algorithm in detail
  • Building the inferred distribution
  • Implementing the Metropolis algorithm in Python

The Metropolis-Hastings Algorithm

    • 1 Videos
    • 1 Plugin
    • 1 Assignment
Show info about module content

1 Assignment

  • MCMC - I

1 Videos

  • The Metropolis-Hastings algorithm

Gibbs Sampling

    • 3 Videos
    • 1 Plugin
Show info about module content

3 Videos

  • Introduction to Gibbs sampling
  • Overview of the Gibbs Sampling algorithm
  • The Gibbs sampling algorithm in detail

Hamiltonian Monte Carlo

    • 2 Videos
    • 2 Readings
Show info about module content

2 Videos

  • Introduction to Hamiltonian Monte Carlo
  • The Hamiltonian Monte Carlo algorithm in detail

2 Readings

  • Hamiltonian Monte Carlo
  • HMC on Stan

Characteristics of MCMC

    • 2 Videos
    • 1 Assignment
Show info about module content

1 Assignment

  • MCMC - II

2 Videos

  • Properties of MCMC - I
  • Properties of MCMC - II

Auto Summary

Embark on a foundational journey into Bayesian Inference with MCMC in this data science and AI course led by Dr. Srijith Rajamohan. Delve into Markov Chain Monte Carlo Methods through engaging Python-based hands-on examples. This 900-minute Coursera course, part of a three-course specialization, utilizes PyMC3 and Jupyter notebooks for practical Bayesian modeling. Ideal for beginners, it offers a starter subscription for accessible learning.

Dr. Srijith Rajamohan
Instructor

Dr. Srijith Rajamohan