DocumentCode
116698
Title
Research of model specifications tests under failure of normality assumption
Author
Volkova, V.M. ; Pankina, V.L.
Author_Institution
Novosibirsk State Tech. Univ., Novosibirsk, Russia
fYear
2014
fDate
2-4 Oct. 2014
Firstpage
552
Lastpage
555
Abstract
Ramsey RESET-test and J-test are designed with the assumption of Gaussian distribution of the random component of the model. It is demonstrated that Ramsey RESET-test is robust to disturbance in the assumption of Gaussian distribution model errors except the distribution with heavy tail. The effects on J-test statistics distribution of model specifications residuals distribution are considered.
Keywords
Gaussian distribution; failure analysis; Gaussian distribution model errors; J-test statistics distribution; Ramsey J-test; Ramsey RESET-test; model specification residual distribution; model specification tests; normality assumption failure; random component; Computational modeling; Computers; Data models; Educational institutions; Gaussian distribution; Noise level; Robustness; J-test; Model specification; Ramsey RESET-test;
fLanguage
English
Publisher
ieee
Conference_Titel
Actual Problems of Electronics Instrument Engineering (APEIE), 2014 12th International Conference on
Conference_Location
Novosibirsk
Print_ISBN
978-1-4799-6019-4
Type
conf
DOI
10.1109/APEIE.2014.7040745
Filename
7040745
Link To Document