On regression method to obtain emission parameters of building materials

Diffusion coefficient and partition coefficient are two critical parameters that describe the emission characteristics of building materials. These parameters could be obtained from least-square regression by fitting the model predictions with the experimental data obtained by widely used small chamber tests. The chamber data quality is important to the reliability of the regression results. In this paper, we study the influence of measurement errors of chamber data as well as data abundance on the regression results of model parameters. The cases studied include one degree, two degrees, and three degrees of freedom. As expected, the more variables to be determined by regression, the larger uncertainties the parameter regression results have. Results also show that prediction of partition coefficient is more likely dependent on the data abundance. The results of this study could be useful for guiding further regression work and establishing the requirements of chamber tests.