Saddlepoint Series for Distribution Functions
Recall from §3 that calculating of distribution function approximations from Edgeworth density approximations was a simple matter. The Edgeworth series for the density is a linear combination of derivatives of the normal distribution function, and hence is easily integrated to give a corresponding cumulative distribution function approximation. This cumulative distribution function approximation inherits many good properties from the density approximation.
KeywordsTail Probability Normal Distribution Function Saddlepoint Approximation Tail Area Cumulant Generate Function
Unable to display preview. Download preview PDF.