MULTIPLE LINEAR REGRESSION-ARTIFICIAL NEURAL NETWORK HYBRID MODEL PREDICTING UPPER FLAMMABILITY LIMIT VOLUME PERCENT OF PURE ORGANIC COMPOUND

摘要

PURPOSE: A hybrid MLR-ANN model for predicting UFVP of pure organic compounds is provided to make ANN outputting the UFVP and to improve prediction performance. CONSTITUTION: A molecular descriptor value for UFVP(upper flammability limit volume percent) of hydrocarbon series compounds is prepared. Experimental data is separated into training and testing groups. The optimal MLRM(multiple linear regression model) for the training group is detected. The optimal ANNM(artificial neural network model) is divided into three groups and detected.