The availability of large environmental datasets and increased computational capability has motivated researchers to propose innovative techniques to mine information from data. The Evolutionary Polynomial Regression (EPR) is a hybrid data-driven technique that combines genetic algorithms and numerical regression for developing easily interpretable mathematical model expressions. EPR is a multi-objective search paradigm for producing multiple models by simultaneously optimizing accuracy and parsimony of resulting expressions. The EPR MOGA-XL is an MS-Excel add-in that allows the user to launch an EPR run as a function in MS-Excel, thereby exploiting a familiar environment to perform data-driven modeling. Inputs and outputs can be easily selected from a spreadsheet, while a separate sheet containing all EPR modeling options can be modified and retrieved for future analyses. The expression(s) of model(s) obtained, the model predictions and fitness indicators are stored in a separate Excel file, allowing subsequent multiple analyses. An application of EPR-MOGA-XL is presented and discussed.
EPR-MOGA-XL: an excel based paradigm to enhance transfer of research achievements on data-driven modeling
LUIGI BERARDI
;
2012-01-01
Abstract
The availability of large environmental datasets and increased computational capability has motivated researchers to propose innovative techniques to mine information from data. The Evolutionary Polynomial Regression (EPR) is a hybrid data-driven technique that combines genetic algorithms and numerical regression for developing easily interpretable mathematical model expressions. EPR is a multi-objective search paradigm for producing multiple models by simultaneously optimizing accuracy and parsimony of resulting expressions. The EPR MOGA-XL is an MS-Excel add-in that allows the user to launch an EPR run as a function in MS-Excel, thereby exploiting a familiar environment to perform data-driven modeling. Inputs and outputs can be easily selected from a spreadsheet, while a separate sheet containing all EPR modeling options can be modified and retrieved for future analyses. The expression(s) of model(s) obtained, the model predictions and fitness indicators are stored in a separate Excel file, allowing subsequent multiple analyses. An application of EPR-MOGA-XL is presented and discussed.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
