Lagrange (2D), piecewise interpolation (2D), Hermite (3D), spline (3D and smooth). Fitting: given a limited number of data points, an approximate function is obtained, which does not require passing through the known data points, but only requires minimizing the total deviation at these points in a certain sense.

The formulation of curve fitting problem is a set of (two-dimensional) data, that is, n points (xi, yi) on the plane, i= 1, 2, n, xi are different from each other. Find a function (curve) y=f(x) to make f(x) closest to all data points under certain criteria, that is, curve fitting.