Symbolic regression is a technique in machine learning that attempts to create succinct equations to describe data. It can be used in physics or engineering to suggest potential physical laws from empirical data. Symbolic regression is named to suggest the connection with linear regression. Whilst linear regression can only create linear relationships such as y = 3x+5, symbolic regression can look for complex formulae such as F = q(Ef + B v sin θ).
Used on Chap. 18: page 448