Classification and Regression Tree Approach for Prediction of Potential Hazards of Urban Airborne Bacteria during Asian Dust Events.

摘要: Despite progress in monitoring and modeling Asian dust (AD) events, real-time public hazard prediction based on biological evidence during AD events remains a challenge. Herein, both classification regression tree (CART) multiple linear (MLR) were applied to assess the applicability of for potential urban airborne bacterial hazards using metagenomic analysis qPCR. In present work, Bacillus cereus was screened as pathogenic candidate positively correlated with PM10 concentration (p < 0.05). Additionally, detection bceT gene qPCR, which codes an enterotoxin B. cereus, significantly increased The CART approach more successfully predicted relatively high coefficient determination (R2) small bias, smallest root mean square error (RMSE) absolute (MAE) compared MLR approach. Regression analyses from model showed that concentration, 78.4 µg/m3 92.2 µg/m3, is important atmospheric parameter affects events. results show may be useful effectively derive predictive understanding thus has possible improving decision-making tools environmental policies associated air pollution health.