Gaussian Process Regression and Classification with Stan

Tentative Agenda

6:00pm - 6:20pm: - Pizza, Drinks, Socialize
6:20pm - 6:30pm: - Installing pre-requisites (R, RStudio (https://www.rstudio.com/products/rstudio/download/preview/), rstan (https://github.com/stan-dev/rstan/wiki/RStan-Getting-Started), bayesplot (https://github.com/stan-dev/bayesplot))
6:30pm - 8:00pm: - Presentation, Demo & Hands-on programming

Abstract:

Gaussian processes (GPs) are a powerful, non-parametric Bayesian method that can be used in supervised learning and classification problems. The main advantages of this method are the ability of GPs to provide uncertainty estimates and to learn functions of the underlying phenomena from training data. The aim of this meetup is to introduce GPs for regression and classification with Stan. We will discuss the basics of GPs as a nonparametric Bayesian method. We will also discuss how to build GP models in computer code using a new exciting programming paradigm called Probabilistic Programming (PP). Particularly we shall use Stan (within R), a PP language, to build GP models for regression and classification.

Pre-requisites:

Please come with a laptop with the following software installed:

• R + RStudio:- Follow this link (https://www.r-project.org/) to install R & this link to install the LATEST version (https://www.rstudio.com/products/rstudio/download/preview/) of RStudio.

• rstan:- Follow this link (https://github.com/stan-dev/rstan/wiki/Installing-RStan-on-Mac-or-Linux) to install rstan.

• bayesplot:- Follow this link (https://github.com/stan-dev/bayesplot) to install bayesplot

Speaker Bio:

Daniel Emaasit (http://www.danielemaasit.com/) is a PhD Candidate of Transportation Engineering at UNLV. His research interests involve developing flexible probabilistic/Bayesian machine learning models (http://mlg.eng.cam.ac.uk/pub/topics/#np) for high-dimensional data with applications to urban mobility, travel demand modeling, and highway safety analysis.