Answers are grounded in this paper and cite their sources.
Within classical constrained approximation, we explicitly construct the unique endpoint-interpolating one-sided quartic minimax polynomials for functions with negative fifth derivative. Their Hermite contact patterns are selected by a unique scalar balance of two error maxima. Although the contact families cross, their lobe maxima have opposite monotonicity; a polynomial-difference certificate verifies optimality over the full feasible class. The exact model constant C=(50−225)/125 gives sharp derivative-based error bounds. For even degree-eight sinc approximation, these yield certified uniform errors and exact shrinking-interval contact limits. A higher-even-degree sufficient optimality certificate is recorded separately, without asserting existence of balanced contacts in every degree. Endpoint-adjusted Taylor bounds improve five specified published Cusa–Huygens estimates and a stronger endpoint-matching lower bound; higher-order comparisons delimit these improvements. Two rational denominator families are also completely classified.
No takes yet. Share an insight, caveat, or question.
Mohammad W. Alomari (2026) studied this question.
Synapse has enriched 10 related papers. Consider them for comparative context: