Equation discovery

Find the equation in your data.

Upload your data. Get a model you can read.

You don't choose the form of the curve: Diafunc searches for it, across all your input columns at once.

Curve fitting and nonlinear regression tune the constants of a form you picked. Here the form itself is discovered, and the result is a few readable lines, scored on rows the search never saw. The method is called Symbolic Regression.

What you get

· A model you can read
· Its quality, measured on rows it never saw
· A full report in plain language
· Your questions, answered
· Ready to use: as code, as an API, as a PDF report

The model Diafunc discovered for the resistance of a yacht hull: one readable line, quality 99.5 percent, well ahead of the linear regression baseline. — Real output. 308 towing-tank measurements of yacht hulls, six inputs. The model was discovered, not written by hand.

Quality 99.5 percent, complexity 28.3 percent, error 0.3 percent. — Real output. 308 towing-tank measurements of yacht hulls, six inputs. The model was discovered, not written by hand.

50,000 free credits to start: about five full analyses. No credit card.

Hosted in the EU Private by default Cancel anytime

Start free

Questions? Read the FAQ