This dataset contains synthetically generated quadratic and cubic polynomial equations along with their exact roots.

The goal of this dataset is to explore whether machine learning models can learn underlying algebraic structures, such as:

• The quadratic formula
• Vieta’s relations
• Complex conjugate symmetry
• Root behavior of cubic equations
  The model should at least predict one root of the equation

Dataset Contents:

• quadratics.csv

  • Coefficients: a, b, c :
  • Roots (real & complex): split into real and imaginary parts

• cubics.csv

  • Coefficients: a, b, c, d
  • Three roots per equation, represented as real and imaginary components

Motivation

Instead of symbolic solvers, this dataset challenges ML models to infer mathematical laws purely from data.

This makes it useful for:

• Regression models
• Neural networks
• Symbolic regression
• ML interpretability research

All values are exact (numerically computed) and generated without noise.