Regression Models for the Field Electron Emission Signal.

Autor: Egorov, N. V., Varayun', M. I., Bure, V. M., Antonov, A. Yu.
Zdroj: Journal of Surface Investigation: X-Ray, Synchrotron & Neutron Techniques; 2020, Vol. 14 Issue 6, p1394-1402, 9p
Abstrakt: The field electron emission signal is described using the regression approach with three models. Each regression function has its own theoretical justification. The three-parameter models make it possible to take into account the deviations of the I‒V characteristics from a straight in the Fowler–Nordheim coordinates in both the high- and low-voltage region. For clear comparison, investigated data samples are obtained in the numerical experiment with specified signal parameters. A special noise model is considered, in which the use of conventional linearization over the parameters seems fair. Attention is paid to checking the regression residuals for normality of their distribution. The parameters of the proposed models are cross estimated and their statistical significance with a confidence level of 95% is verified. It is shown that, for each set of experimental data, a separate model may be necessary. The adjusted coefficient of determination is used as a measure of the dependence of the response on the factor. [ABSTRACT FROM AUTHOR]
Databáze: Complementary Index