A class of cause-and-effect models, where the first relationship between cause (independent variable) and effect (dependent variable) is determined, which then is used to prepare a forecast. If the model states that sales are a function of advertising, then the model will first compute the relationship between sales and advertisements, using the historical data, and then use that relationship to arrive at a forecast.

Previous Entry
Reference Forecast
Next Entry
Regression Coefļ¬cient