https://github.com/aybchan/gaussianprocess

interactive gaussian process modelling with d3.js
https://github.com/aybchan/gaussianprocess

bayesian-statistics d3 d3js data-visualization gaussian-process-regression gaussian-processes

Last synced: 6 months ago
JSON representation

interactive gaussian process modelling with d3.js

• Host: GitHub
• URL: https://github.com/aybchan/gaussianprocess
• Owner: aybchan
• License: mit
• Created: 2018-04-02T20:23:54.000Z (almost 8 years ago)
• Default Branch: master
• Last Pushed: 2018-04-09T11:16:52.000Z (almost 8 years ago)
• Last Synced: 2025-06-06T12:49:28.726Z (8 months ago)
• Topics: bayesian-statistics, d3, d3js, data-visualization, gaussian-process-regression, gaussian-processes
• Language: JavaScript
• Homepage: https://gaussianprocess.herokuapp.com/
• Size: 315 KB
• Stars: 5
• Watchers: 1
• Forks: 0
• Open Issues: 0
• Metadata Files:
  • Readme: README.md
  • License: LICENSE

Awesome Lists containing this project

README

          
# Gaussian process interactive playground

![Bacchus (2006) — Cy Twombly](twombly.jpg)

An interactive demo of a simple gaussian process regression model implemented with d3.js.

## Installation instructions

1. Requires [node.js](https://nodejs.org) to run

2. In the terminal do:

```

git clone https://github.com/aybchan/gaussianprocess

cd gaussianprocess

npm install

npm start

```

3. Open http://localhost:3000 in your browser

## Playground instructions

The canvas is populated with draws from the gaussian process prior distribution (green), some initial observations (blue) and the gaussian process posterior (red) conditioned on this data. Click on the canvas to add a new observation, this will appear as a cyan dot.the GP posterior samples will update to reflect the incorporation of new observations in the GP model. The GP posterior mean and the band of uncertainty (posterior mean ± 2σ—the black lines) will also update.

The configuration of the playground can be changed in public/javascripts/app.js.