Prediction regression provides the automatic identification of predictor variables through multiple regression analysis and advanced statistical tests. Regression results are presented in a simple and easy to understand format to quantify the relative influence of each input variable. The work flow facilitates and iterative process to test, maintain and discard variables until a prediction regression equation can be established with maximum confidence. Supplementary statistical analysis to reveal underlying data relationships include autocorrelation under the Dubin-Watson statistic and multicollinearity between individual independent variables. Standard tests include F statistic confidence intervals, adjusted R-squared, standard errors, t-test statistics and p values. The resulting prediction regression equation can subsequently applied to integrated forecasting methods or custom data for the independent variables to produce predictions and forecasts of desired period length. Built-in forecasting options for predictive analysis include linear, polynomial and exponential methodologies. Prediction regression runs in Excel for both Mac and Windows platforms.
Software by Business Spreadsheetswww.business-spreadsheets.com