For every real , the mean value theorem bounds on . Hence the apparent singularity at zero cancels, and the remaining bound is Lebesgue integrable. On a bounded parameter interval , every positive-order parameter derivative of the integrand is bounded in absolute value by . Repeated dominated differentiation under the integral sign therefore proves that is a smooth function, with
In particular, , , and . Thus is strictly convex, lies above its tangent at zero, and satisfies for and for . The inequalities follow directly as well by comparing to one. The substitution removes the square-root endpoint singularity for numerical integration:
Figure 1.
Convexity and derivative bounds for the exponential endpoint integral
. The derivative is below two for negative parameters and above two for positive parameters. The integral lies above its tangent at zero. Substitution removes the endpoint singularity before numerical quadrature. Details.

Articles by others on the same topic (0)

There are currently no matching articles.