https://github.com/sinclairzx81/neuron-function-approximation
An experiment using neural networks to approximate pure functions
https://github.com/sinclairzx81/neuron-function-approximation
experiments machine-learning neural-networks
Last synced: 26 days ago
JSON representation
An experiment using neural networks to approximate pure functions
- Host: GitHub
- URL: https://github.com/sinclairzx81/neuron-function-approximation
- Owner: sinclairzx81
- License: other
- Created: 2017-09-23T08:38:47.000Z (almost 9 years ago)
- Default Branch: master
- Last Pushed: 2017-09-23T10:03:00.000Z (almost 9 years ago)
- Last Synced: 2024-10-19T11:12:36.086Z (over 1 year ago)
- Topics: experiments, machine-learning, neural-networks
- Language: TypeScript
- Size: 27.3 KB
- Stars: 1
- Watchers: 3
- Forks: 0
- Open Issues: 0
-
Metadata Files:
- Readme: readme.md
- License: license.md
Awesome Lists containing this project
README
# neuron: function approximation
An experiment using neural networks to approximate pure functions [demo](https://sinclairzx81.github.io/neuron-function-approximation/)
### overview
This project is a experiment to approximate pure functions using neural networks. The project
provides a small interactive scripting interface allowing end users to script single parameter
functions ```x' = f(x)``` and a secondary scripting interface for tweaking the network
parameters. The function ```f(x)``` output is clamped between a -1 and 1 range with the network
activation set to ```tanh``` end to end.
The goal of this project is to test how well neuron can approximate the given function ```f(x)``` by
observing its input and output where input ```x``` is also between a -1 to 1 range.
### building the project
clone this repository and run the following. This will start a small http server on port 5000.
```
npm install
npm start
```