Statistical Models of Characteristics of Metal Vapor Lasers

Bok av Snezhana Georgieva Gocheva-Ilieva
In this book, various multivariate statistical techniques are applied in order to assess and predict the experiment. There are developed multiple linear and quasilinear regression and other parametric and non-parametric models for modelling laser efficiency and output power. The best models are obtained using non-parametric method of multivariate adaptive regression splines which can account for the local changes in the behaviour of the investigated quantities and possess powerful predictive capabilities. The obtained results can be considered a first attempt and foundation for the statistical processing of the accumulated experimental data for guiding future experiment in the field of lasers