Paper Title
Cheminformatics Methods for Rapid Prediction of Physicochemical Properties

Abstract
In this project was developed cheminformatics methods based on machine learning for a quick preview correlation of energy with the size of organic compounds, the enthalpy of combustion of hydrocarbons and statistical separation of compounds. The data set of organic substances calculated by quantum chemistry was built. However, despite the theoretical chemical calculations based on Density Functional Theory (DFT) to provide estimates of various properties with increasing accuracy, its computational cost is relatively high for many situations. The correlation with the enthalpy of combustion allows the calculations of engineering and unit operations are faster and more accurate. In this work the correlation obtained was R = 1 for combustion enthalpy. Application of Principal Component Analysis and Dendrogram were used to separate alcohol of hydrocarbons by recognizing patterns using vibrational properties calculated by DFT, being a useful tool for physical-chemical analysis. Keywords� Density Function Theory, Enthalpy, Organic Compound, Principal Component Analysis, Dedrogram