Polynomial degree reduction in the discrete \(L_2\)-norm equals best Euclidean approximation of \(h\)-Bézier coefficients (Q285260)

Considering the classical \(L_2\)-inner product and the Euclidean inner product of the Bézier coefficients on the linear space \(\mathbb P_n\) of polynomials of degree at most \(n\), \textit{D. Lutterkort} et al. [Comput. Aided Geom. Des. 16, No. 7, 607--612 (1999; Zbl 0993.41016)] proved that the best degree reduction of a given polynomial \(P\) from degree \(n\) to \(m\) with respect to the norms induced by the two inner products are equivalent (cf. [\textit{Y. J. Ahn} et al., Comput. Aided Geom. Des. 21, No. 2, 181--191 (2004; Zbl 1069.41504)]). Introducing a sample parameter \(h\), the author provides a discrete analogue to the foregoing result. As \(h\) tends to zero, the above result of Lutterkort et al. [loc. cit.] is recovered. Adequacy of using the \(h\)-Bernstein bases to approach the problem of polynomial degree reduction with respect to weighted discrete norms is also highlighted.