The coefficient of determination is, in statistics, an indicator which measures the quality of a prediction. The closer this indicator is to 1, the closer the model is to reality.
coefficient of determination = 1 - Sum of the deviations of the measurement with the squared prediction / Sum of the deviations of the measurement with the squared mean
where R² is the coefficient of determination, y "i" the observed measurement, y "hat i" the predicted value and y "bars" the mean
For the following time series:
So the R² is worth: 1- ((50² + 30² + 200²) / ((1500-1510) ² + (1230-1500) ² + (1800-1500) ²)) = 0.73
The predictive model is correct.
The coefficient of determination is often called R² and its English translation is "Coefficient of determination".
Sales & Marketing Advanced subscription