Testing Curve Fit.
Final technical rept.,
ARMY ARMAMENT RESEARCH DEVELOPMENT AND ENGINEERING CENTER WATERVLIET NY CLOSE COMBAT ARMAMENTS CENTER
Pagination or Media Count:
A test is derived for the hypothesis of residual randomness in a curve fit. The test is based on a binomial process where the main problem involved is equivalent to finding the distribution of the number of runs in a sequence of coin tosses. Although this distribution turns out to be quite simple in form and easy to apply, its use in testing curve fit seems to have been neglected in the statistical literature. We also obtain the more general distribution of the number of runs in a sequence of throws of a multifaceted die, which is used to test sequences for randomness per se. Author Keywords Nonparametric Test Hypothesis Testing Residual Analysis Trees Recursions Moments Asymptotic Normality.
- Statistics and Probability