Mean Squared Error (MSE)
This metric is calculated by averaging the square of the errors between the actual values and forecasts for each unit of time. A lower MSE value implies a higher accuracy of the model. This metric is sensitive towards outliers. A drawback of MSE is that it is not scaled in units of the variable of interest (squared error), leading to a metric that cannot be related to the original time series value.
Last modified: Friday May 12, 2023