Evaluation of Fire Risk Based on Generalized Regression
Qian Ding, Zhenghong Peng, Tianzhen Liu, Qiaohui Tong · 2014
Based on generalized regression without known dependent variable, that is, without training samples, an evaluation model of the level of fire hazard is proposed, and a definite linear algorithm without iteration is given. The model realizes regression with latent variable using reasonable prescription constraint. The data example is given by means of convenient computation software. The method is also suitable to other evaluation problems with similar data constructions.