The quadratic form of non-central normal variables is presented based on a sum of weighted independent non-central chi-square variables. This presentation provides moments of quadratic form. The maximum entropy method is used to estimate the density function because distribution moments of quadratic forms are known. A Euclidean distance is proposed to select an appropriate maximum entropy density function. In order to compare with other methods some numerical examples were evaluated. Also, for discrimination between two groups by the Euclidean distances, we obtained a stochastic representation for the linear discriminant function using the quadratic form. The maximum entropy estimation was an acceptable method to approximate the distribution of quadratic forms in normal variables.
Rekabdar, G., & Chinipardaz, R. (2017). Approximating the Distribution of Indefinite Quadratic Forms in Normal Variables by Maximum Entropy Density Estimation. Journal of Modern Applied Statistical Methods, 16(2), 359-377. doi: 10.22237/jmasm/1509495540