US Patent:
20120095800, Apr 19, 2012
Inventors:
SAEED BAGHERI - CROTON ON HUDSON NY, US
DOUGLAS C. BUNCH - RICHARDSON TX, US
THERESA E. ELL - BRONXVILLE NY, US
KRISHNA C. RATAKONDA - YORKTOWN HEIGHTS NY, US
KATRINA M. REFFETT - GREENUP KY, US
ROSE M. WILLIAMS - WAPPINGER FALLS NY, US
Assignee:
INTERNATIONAL BUSINESS MACHINES CORPORATION - ARMONK NY
International Classification:
G06Q 10/00
Abstract:
A computer implemented method, system, and/or computer program product predicts the profitability of a current project. Historical data from the current project represents profit data points at predetermined intervals of time during the current project. If a linear regression model of the historical data does not adequately describe the historical data, then a polynomial regression model to describe the historical data is generated. If the polynomial model does not adequately describe the historical data, then another project that is similar to the current project is located. This other project has its own set of historical data. If the historical data from the other project fits the polynomial regression model created from the historical data of the current project, then the polynomial regression model is trusted to accurately predict the profitability of the current project, even though the current project's historical data did not adequately fit the polynomial regression model.