Full text available in Adobe Acrobat format
Volume (52) Number 2 pp. 75-84
Andrzej Kornacki 1, Andrzej Bochniak 1
1Department of Applied Mathematics and Computer Science, University of Life Sciences in Lublin, Akademicka 15, 20-950Lublin, Poland
The use of outlier detection methods in the log-normal distribution for the identification of atypical varietal experiments
Abstract
In this study the Akaike information criterion for detecting outliers in a log-normal distribution is used. Theoretical results were applied to the identification of atypical varietal trials. This is an alternative to the tolerance interval method. Detection of outliers with the help of the Akaike information criterion represents an alternative to the method of testing hypotheses. This approach does not depend on the level of significance adopted by the investigator. It also does not lead to the masking effect of outliers.
Keywords: outliers, log-normal distribution, atypical variety trials, hypothesis testing, masking of outliers, wheat, entropy
DOI: 10.1515/bile-2015-0007
For citation:
| MLA | Kornacki, Andrzej, and Andrzej Bochniak. "The use of outlier detection methods in the log-normal distribution for the identification of atypical varietal experiments." Biometrical Letters 52.2 (2015): 75-84. DOI: 10.1515/bile-2015-0007 |
| APA | Kornacki, A., & Bochniak, A. (2015). The use of outlier detection methods in the log-normal distribution for the identification of atypical varietal experiments. Biometrical Letters 52(2), 75-84 DOI: 10.1515/bile-2015-0007 |
| ISO 690 | KORNACKI, Andrzej, BOCHNIAK, Andrzej. The use of outlier detection methods in the log-normal distribution for the identification of atypical varietal experiments. Biometrical Letters, 2015, 52.2: 75-84. DOI: 10.1515/bile-2015-0007 |