panzer/polynomial-regression

Fit 2D points to a n-degree polynomial function using TensorFlow.js. Visualization with p5.js

Polynomial Regression

Live Demo

Click here to play around with it on your own!

What is it?

Paraphrasing from Wikipedia, "polynomial regression involves modeling the relationship between the variables x and y with an nth degree polynomial function."

A polynomial function takes the form

This program takes a set of (x,y) points as an input, and attempts to produce a polynomial that fits those points. This is known as curve fitting. The program does this by gradually adjusting the coefficients-- the a values.

Inspiration

Daniel Shiffman's Coding Train Youtube channel. Specifically, this video.

I took this as an opportunity to get started with basic machine learning tasks. With the recent release of TensorFlow.js and my growing Javascript experience, I figured now was a perfect time to start tinkering!

Contributing

Feel free to clone the repo and play around! The sketch.js file contains all of the core logic and is fairly well documented.

For any bugs/unexpected behavior/suggestions, mark it as an Issue on GitHub and I'll fix it ASAP! Or, even better, submit a Pull Request.

License

All original work within this repository is available under the MIT License. See the LICENSE file for more information.