Andrew Ribeiro - Andrew@kexp.io
Talk given July 2018 to Danbury AI
Often when doing research in machine learning you need to visualize your data and results to gain insights. In addition,publications like https://distill.pub show us how much value well put together visualizations can add to the exposition of machine learning research. In this talk we will introduce the basics of a popular library for creating robust web-based data visualizations: D3.js.
Embedded_Template
A basic project template.
This project does not require node.js or knowledge of modern JavaScript development practices. You should start here if you are a JavaScript novice.
PseudoShop
Produces a table of data sampled from a hard coded probability distribution. This project demonstrates the basic principles of using D3.js bind data to the DOM and perform data-driven transformations.
StockSim
A stock simulator. Shows you how to make line graphs for time series data.
Webpack_Template
This folder contains a webpack template project. You can copy this folder and use it as a base for your project. It was used as the template project for the following projects.
CSV_Viewer
Drag and drop a CSV file into the browser and have d3 read it in transform the data to HTML.
MNIST
Interactively view 55,000 MNIST digits. This project shows how image data can be rendered with D3 in the SVG canvas, how to make animations, and how to use input elements to interact with your visualization.