Genetic

Genetic algorithm

Curve fitting is a well researched problem. There are few approaches that use genetic algorithms for curve fitting. We try a new approach to curve fitting where we don't focus on performance but rather try to create "real" equations. This differs from the traditional approach, where parameters of a harmonic function are varied to produce a specific curve.

The visualization is written in D3 and javascript; however, the server and genetic algorithm are written entirely in clojure.

Examples

Generated by code

For goal graph - blue is randomly generated, red is our match.

Generated by code

For goal graph - blue is randomly generated, red is our match.

Setup and run

lein run -m myapp.core

Then navigate your browser to localhost:5000 to see some graphs