Abstract—In this paper, I propose a fitting algorithm for a probability distribution for observations by using the Kolmogorov-Smirnov test. Drezner et al. propose an algorithm that calculates the mean and standard deviation of a normal distribution of observations in order to minimize the KS statistic of the Kolmogorov-Smirnov test. I generalize the algorithm of Drezner et al. and obtain the necessary conditions that enable other probability distributions to be applied. I show that it may be applied to well-known probability distributions.
Index Terms—Closest fit, kolmogorov-smirnov test, probabilistic distribution.
N. Sakamoto is with the Department of Information and Communication Engineering, Tokyo Denki University, Tokyo, Japan (e-mail: sakamoto@c.dendai.ac.jp).
[PDF]
Cite:Naoshi Sakamoto, "A Generalized Fitting Algorithm Using the Kolmogorov-Smirnov Test," International Journal of Computer Theory and Engineering vol. 9, no. 2, pp. 142-146, 2017.
