Neural Network Demo

This repository contains the neural network and perceptron demonstration program for the lecture “Kognitive Systeme” at the Interactive Systems Lab (ISL) at the Karlsruhe Institute of Technology.

This demo runs completely client-side in the browser. The index.html file in the gh-pages branch can be opened on a local webserver (python3 -m http.server). A hosted version is available here.

Development

This project is written in TypeScript, a statically typed superset of JavaScript.

I recommend Visual Studio Code (crossplatform and opensource) for development. An alternative is Atom + atom-typescript.

React is used with JSX for GUI state handling.

The compiled files are included in the gh-pages branch

Setup

git checkout gh-pages # to set upstream
git checkout master
git worktree add bin gh-pages
yarn install

Then build via $ webpack.

Use

python3 -m http.server &
yarn run webpack --watch &

for automatic compiling and deploying to http://localhost:8000/bin.

Use

NODE_ENV=production webpack

for an optimized build.

Use Simulation.instance to access the main object and all it's children.

Use cd bin; git add -A; git commit -m'update binaries'; git push to update the binaries.