Powerful Goodness-of-fit Tests Based on the Likelihood Ratio
Jin Zhang
Abstract
Jin Zhang
Abstract
Summary A new approach of parameterization is proposed to construct a general goodness-of-fit test. It can not only generate traditional tests (including the Kolmogorov–Smirnov, Cramér–von Mises and Anderson–Darling tests) but also produce new types of omnibus tests, which are generally much more powerful than the old ones.
OpenAlex reports 229 citations for this work. Citation counts describe recorded attention and do not establish research quality.
A contribution statement is not available in the OpenAlex record.
Method details are not available in the OpenAlex metadata.
Findings are not separately available in the OpenAlex metadata.
Limitations are not available in the OpenAlex metadata.
Application details are not available in the OpenAlex metadata.
Summary A new approach of parameterization is proposed to construct a general goodness-of-fit test. It can not only generate traditional tests (including the Kolmogorov–Smirnov, Cramér–von Mises and Anderson–Darling tests) but also produce new types of omnibus tests, which are generally much more powerful than the old ones.
Key concepts: Goodness of fit, Anderson–Darling test, Omnibus test, Mathematics, Statistics, Construct (python library), Kolmogorov–Smirnov test, Statistical hypothesis testing