Feyn - Symbolic AI using the QLattice

The QLattice is an explainable symbolic AI that accelerates research and scientific discovery.

With Feyn, exploring and understanding relationships in your data is made easy, putting the science in data science.

Quick start

From zero to scientific hero.
Just pip install feyn.

Feyn Quick Start

Using Feyn

Get started with complete examples for classification and regression.

Feyn in 5 minutes

Learn by example

Visit our learning page for ideas on where to get started.

Learn

Extra, extra!

Feyn 3.4.0 is out!

This update brings automatic detection of model output, style improvements to make some plots more legible, bugfixes and further calibration of the automatic type detection.

See the full changelog here

In other news...

Read about our other recent new features and additions to the Feyn Documentation.
You can also check out our changelog here.

Understand your models

Understand and compare your models more easily with the automatic response plot.

Insert data -
press play!

Get started even faster with automatic type detection for your data when using Auto Run.

Stratify your data

Stratify your data splits across multiple columns with the built-in function.

Get going with our Beginner Tutorials for common cases

Learn the QLattice workflow through solving well-known problems.

Titanic Dataset

Explore the Titanic dataset using Feyn and the QLattice for classification.

AirBnB Rental Prices

Explore the AirBnB dataset, using Feyn and the QLattice for regression.

Pulsar Stars

A fun classification problem for classifying whether a star is a pulsar.

Dust off your QLattice skills in life sciences

Explore how you can use the QLattice to solve problems in life science.

Stanford COVID-19 vaccine

Explore the Stanford COVID-19 dataset, using the QLattice for regression.

Feyn and R for ASOs

Explore antisense oligonucleotides in your familiar workflow combining Feyn with R.

Using R and Feyn for antisense olignucleotides

Detecting Liver Cancer (HCC)

Determine liver cancer in patients using biomarkers