Measuring prediction errors in time series forecasting
Measuring prediction errors is an important step in the process of developing a predictive machine learning model. In time series forecasting, model performance is typically measured with different error metrics, each of which having own advantages and disadvantages. Therefore, the different metrics are typically used in combination. This blog post presents a cheat sheet containing six of the most common error metrics used in time series forecasting. The cheat sheet contains the mathematical formula for these metrics, a short code sequence to implement them in Python, and some hints for their interpretation.
Metrics for measuring prediction errors
The performance of time series forecasting models is measures by the deviations between the predictions (y_pred) and the actual values (y_test). If the prediction is below the actual value, the prediction error is positive. If the prediction lies above the actual value, the prediction error is negative.
The following six metrics that are commonly used to measure prediction errors of time series forecasting models:
- Mean Absolute Error (MAE)
- Mean Absolute Percentage Error (MAPE)
- Median Absolute Error (MedAE)
- Mean Squared Error (MSE)
- Root Mean Squared Error (RMSE)
- Median Absolute Percent Error (MdAPE)
You may wonder why there are multiple error metrics. The reason is that each metric by itself can only cover a part of the overall picture. For instance, imagine you have developed a model to predict the consumption of a power plant. The predictions of the model are generally accurate, but in few cases the predictions are very wrong. In other words, there outliers among the prediction errors. To understand this situation, it is not sufficient to calculate the average prediction error. In order to identify outliers, you will need to combine different error metrics. The following section introduces the six error metrics.
Mean Absolute Error (MAE)
Mean Absolute Error (MAE) is a metric that is commonly used to measure the arithmetic average of deviations between predictions and actual values. An MAE of “5” tells us that on average our predictions deviate from the actual values by 5. Whether this error is considered small or large will depend on the application case and the scale of the predictions. For instance, 5 nanometers in the case of a building might be small, but if it’s five nanometers in the case of a biological membrane, it might be large. So when working with the MAE, mind the scale.
- Since it is calculated on absolute values, positive and negative deviations from the actual value are taken into account equally in the calculation.
- The MAE is sensitive to outliers, as large values can have a strong impact. For this reason, the MAE should be used in combination with additional metrics.
- The MAE shares the same unit with the predictions.
Mean Absolute Percentage Error (MAPE)
The mean absolute percentage error calculates the mean percentage deviation between predictions and actual values.
- The mean absolute percentage error is scale independent, making it easier to interpret.
- Not to be used whenever a single value is zero
- Puts a heavier penalty on negative errors
Median Absolute Error (MedAE)
The Median Absolute Error (MedAE) calculates the median deviation between predictions and actual values.
- The MedAE has the same unit as the predictions.
- A MedAE of value 10 means that 50% of the errors are greater than 10 and 50% of the errors are smaller.
- The MedAE is resistant to outliers. It is therefore often used in combination with the MAE. A strong deviation between MAE and MedAE is an indication that there are outliers in the errors where the prediction model deviates more from the actual value than on average.
Mean Squared Error (MSE)
We can calculate the MSE by measuring the average squares of the differences between the estimated values and the actual values. The formula of the MSE is:
- Since all values are squared, the MSE is very sensitive to outliers.
- An MSE that is much larger than the MAE indicates strong outliers among the prediction errors.
Root Mean Squared Error (RMSE)
The root mean squared error is another standard way to measure the performance of a forecasting model.
- Has the same unit as the predictions
- Good measure of how accurately the model predicts the response
- Robust to outliers
Median Absolute Percentage Error (MdAPE)
- Not to be used whenever a single value is zero
- More robust to distortion from outliers than the MAPE
This post has presented an error metrics cheat sheet with six metrics for measuring prediction errors in time series forecasting. In addition, this post has presented each error metric along with the particularities of their application.
Leave a comment if you have remaining questions or want remarks.
I have created another blog post, in which I demonstrate how to use the metrics.