Analysis on Difference of Contaminated Normal Distribution PDF

Abstract:

As a very important distribution, contaminated normal distribution play a great role in data processing. The probability density function (PDF) feature of the contaminated normal distribution was investigated. The Kullback-Leibler distance is suggested for measuring PDF difference between mean shift model and variance inflation model. Numerical calculations show that the PDF difference of two kinds of model is related to mean shift parameter λ and the variance inflation factor α closely when the main distribution is the standard normal distribution and the relationship is nonlinear proportional.

• Periodical: Applied Mechanics and Materials (Volumes 409-410)
• Pages: 1661-1666
• DOI: https://doi.org/10.4028/www.scientific.net/AMM.409-410.1661
• Online since: September 2013
• Authors: Fang Bin Zhou, Yun Kai Guo
• Keywords: Contaminated Distribution Model, Contaminated Normal Distribution, Kullback-Leibler Distance, PDF

References

[1] Huber P J, Robust Statistics, New York, Wiley, (1981).

[2] Jiangwen ZHOU, Classical theory of error and robust estimation, Acta Geodaetica et Cartographica Sinica18 (1989)115-120.

[3] Tukey J W, A survey of sampling from contaminated distribution, in: contribution to probability and statistics. Stanford, Stanford University Press, (1960).

[4] Staudte R G, Sheather S J, Robust estimation and testing, New York, Wiley, (1990).

[5] Shijian ZHOU, Yongqi Chen, Kangwei DENG, The robust measure of Lp estimation under contaminated distribution, Journal of Wuhan Technical University of Surveying and Mapping24 (1999)71-74, 89.

[6] Zhizhong WANG, Jianjun ZHU, Choice of estimation of unknown parameter under contaminated error model, Trans. Nonferrous Met. Soc. China9 (1999)852-856.

[7] Jianjun ZHU, The Theory of errors and surveying adjustment under contaminated error model, Changsha, Central South University, (1998).

[8] Liyan WANG, Yi ZHANG, Enmin FENG, An estimation method of contamination distribution density function, Journal of Dalian University of Technology43 (2003)551-554.

[9] Xiaohan YANG, Non-parameter estimation of contaminated data, Journal of Tongji University29 (2001)700-702.

[10] Yuanxi YANG, Hongzhou CHAI, Lijie SONG, Approximation for contaminated distribution and its applications, Acta Geodaetica et Cartographica Sinica28 (1999)209-214.

[11] Shijian ZHOU, Shaobing ZENG, The Equivalence of Mean Shift Model and Variance Inflation Model, Mine Surveying1 (1995)8-10.

[12] Fangbin Zhou, Jianjun ZHU, Yongqi Chen, Approximate estimation of the p-norm distribution entropy, Bulletin of Surveying and Mapping12(2012) 23-24, 36.

[13] Zelin CAI, Kaican LI, Maximum Kullback-Leibler Distance of Some Conventional Distributions, Journal of Wuhan University ( Nat. Sci. Ed. ) 53 (2007)513-517.