Abstract
Continuous goodness-of-fit (GOF) is a classical hypothesis testing problem in
statistics. Despite numerous suggested methods, the Kolmogorov-Smirnov (KS)
test is, by far, the most popular GOF test used in practice. Unfortunately, it
lacks power at the tails, which is important in many practical applications.
In this paper we make two main contributions: First, we propose the \em
Calibrated KS (CKS) statistic, a novel test statistic based on the following
principle: Instead of looking for the largest deviation between the empirical
and assumed distributions, as in the original KS test, we search for the
deviation which is most statistically significant. By construction, the
resulting CKS test achieves good detection power throughout the entire range of
the distribution. Second, we derive a novel computationally efficient method to
calculate $p$-values for a broad family of one-sided GOF tests, including the
one-sided version of CKS, as well as the Higher Criticism and the Berk-Jones
tests. We conclude with some simulation results comparing the power of CKS to
several other tests.
Description
[1311.3190] The Calibrated Kolmogorov-Smirnov Test
Links and resources
Tags