Predicting PM 2 . 5 in well-mixed indoor air for a large 2 office building using regression and artificial neural 3 network models 4

10 Although the exposure to PM2.5 has serious health implications, indoor PM2.5 monitoring is not 11 a widely applied practice. Regulations on indoor PM2.5 level and measurement schemes are not 12 well-established. Compared to other indoor settings, PM2.5 prediction models for large office 13 buildings are particularly lacking. In response to these challenges, statistical models were 14 developed in this paper to predict the PM2.5 concentration in well-mixed indoor air in a commercial 15 office building. The performance of different modeling methods, including multiple linear 16 regression (MLR), partial least squares regression (PLS), distributed lag model (DLM), least 17 absolute shrinkage selector operator (LASSO), simple artificial neural networks (ANN), and longPage 2 of 28