Psicothema · 2013 · 359 citations · 34 references
ComputingMeteorologyForecastingMachine LearningTime Series AnalysisForecast Error AnalysisForecast EvaluationForecasting MethodsForecasting UncertaintyTime Series ForecastingTime Series PredictionArima ModelsAutoregressive Integrated Moving AverageForecasting AccuracyForecasting Accuracy EvaluationGoodness-of-fitGoodness-of-fit AssessmentGoodness-of-fit IndicesTime Series Forecast Error
BACKGROUND: The mean absolute percentage error (MAPE) is probably the most widely used goodness-of-fit measure. However, it does not meet the validity criterion due to the fact that the distribution of the absolute percentage errors is usually skewed to the right, with the presence of outlier values. In these cases, MAPE overstates the corresponding population parameter. In this study, we propose an alternative index, called Resistant MAPE or R-MAPE based on the calculation of the Huber M-estimator, which allows overcoming the aforementioned limitation. METHOD: The results derived from the application of Artificial Neural Network (ANN) and Autoregressive Integrated Moving Average (ARIMA) models are used to forecast a time series. RESULTS: The arithmetic mean, MAPE, overstates the corresponding population parameter, unlike R-MAPE, on a set of error distributions with a statistically significant right skew, as well as outlier values. CONCLUSIONS: Our results suggest that R-MAPE represents a suitable alternative measure of forecast accuracy, due to the fact that it provides a valid assessment of forecast accuracy compared to MAPE.
34
Time Series Analysis: Forecasting and Control
Michael D. Geurts, George E. P. Box, Gwilym M. Jenkins · Journal of Marketing Research · 1977 · 19.3K citations
Computing, Data Science, Forecasting +11
F. N. David, John W. Tukey · Biometrics · 1977 · 12.9K citations
Multilayer feedforward networks are universal approximators
HornikK., StinchcombeM., WhiteH. · Neural Networks · 1989 · 9.4K citations