Algorithm 1 Fitting a function via RANSAC

Let f : ℝn → ℝn be the function we wish to fit. Furthermore, let $S={\{({\mathbf{x}}_k,{\mathbf{y}}_k)\}}_{k=1}^m$, such that xk ∈ ℝn corresponds to yk ∈ ℝn. N and ε are parameters.

I* ← ø

for i = 1, …, N do

Fit f to a randomly drawn subset of S

I ← {(x, y) ∈ S : ‖f(x) – y‖ < ε}

if |I| > |I*| then

I* ← I

end if

end for

Fit f to I*