/Bayesian-Statistics-Econometrics

Bayesian Statistics-Econometrics

Primary LanguageJupyter NotebookMIT LicenseMIT

Cover_Github_Repositories

Bayesian Statistics and Econometrics

This is a training session of Bayesian statistical methods, the content presented are essential elements of machine learning framework. The session is prepared for senior quantitative analysts/researchers in hedge fund or other research institutes who wants to refresh Bayesian methods quickly, also perfect for grad student who are interested in quantitative methods in industry. All proprietary data and cases are censored, thus no institutional information or data are revealed in these training materials.

Prerequisites

The courses are not for beginners, the attendees must have working knowledge of linear algrebra, statistics and probability theory, and ideally advanced econometrics skills too.

And also the attendees are assumed to have constant exposure of

  • Python
  • NumPy
  • Matplotlib
  • Statsmodels
  • Pandas

If you are not familiar with linear regression mechanism, take a look at these notes first.

Contents

Advanced Econometric and Statistical Methods

Chapter 1 - Geometry of Odinary Least Squares
Chapter 2 - Statistical Properties of OLS
Chapter 3b - Hypothesis Test and Confidence Interval

It is advised that you download all material and browse in your own computer, since nbviewer has persistent LaTeX rendering errors.

Bayesian Methods

Chapter 1 - Introduction to Bayesian Methods
Chapter 2 - Bayesian Conjugates
Chapter 3 - Bayesian Simple Linear Regression
Chapter 4 - Markov Chain Monte Carlo
Chapter 5 - Metropolis-Hastings Algorithm
Chapter 6 - Gibbs Sampler
Chapter 7 - Revisit Linear Regression

Screen Captures

bayes1 bayes2 bayes3 bayes4 bayes5 bayes6 bayes7