WRI’s patent-pending ExpliFit© software program lets correlates between any combination of spectral, chemical or physical data measurements. For example, it helps “decode” spectra and reveals the link to physical data, connecting master curves to infrared spectra or asphalt SAR-AD fractions. The correlation equations can characterize asphalt rheology using only chemical data, to monitor binder changes in asphalt pavement or shingles, determine surface treatment schedules and milling depths.
Unlike other chemometric regression methods (e.g. Partial Least Squares, Artificial Intelligence, and Neural Networks), ExpliFit© clearly defines the variables of importance. Correlations are presented as measured variables in closed-form equations (no complicated latent variables), and the regression equation can be transported outside the program and easily fit to additional data. For statistical validation, independent variables and the numbers of observations are shown in ANOVA tables. Chemical data, which is often faster and more economical to acquire, can now replace greater time and material consuming as well as more expensive physical data collection methods used for asphalt material development.
ExpliFit© uses Microsoft Excel® files for input and output and provides clear, easy-to-read graphics for fit analyses. Users are guided through the process to prevent overfitting data. ExpliFit© can be used in any discipline where chemical or physical data are generated and where relationships between multiple influences are desired. It has a wide range of applications from biochemistry to petroleum product analysis.