Using the R-MAPE index as a resistant measure of forecast accuracy.

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.

[1]  Milton S. Boyd,et al.  Designing a neural network for forecasting financial and economic time series , 1996, Neurocomputing.

[2]  G. Lewicki,et al.  Approximation by Superpositions of a Sigmoidal Function , 2003 .

[3]  L. Carter,et al.  The role of population size in the determination and prediction of population forecast errors: An evaluation using conifidence intervals for subcounty areas , 1998 .

[4]  Rob Law,et al.  A neural network model to forecast Japanese demand for travel to Hong Kong , 1999 .

[5]  S K Smith,et al.  The relationship between the length of the base period and population forecast errors. , 1990, Journal of the American Statistical Association.

[6]  J. J. Montaño,et al.  Sensitivity Analysis Applied to Artificial Neural Networks for Forecasting Time Series , 2008 .

[7]  P. Rochon,et al.  Effect of regulatory warnings on antipsychotic prescription rates among elderly patients with dementia: a population-based time-series analysis , 2008, Canadian Medical Association Journal.

[8]  J. J. Montaño Moreno,et al.  Artificial neural networks applied to forecasting time series. , 2011, Psicothema.

[9]  P. Goodwin,et al.  On the asymmetry of the symmetric MAPE , 1999 .

[10]  Gwilym M. Jenkins,et al.  Time series analysis, forecasting and control , 1971 .

[11]  Edward G. Carmines,et al.  Reliability and Validity Assessment , 1979 .

[12]  Lyle H. Ungar,et al.  Neural network forecasting of short, noisy time series , 1992 .

[13]  Fabian Ramseyer,et al.  Modeling psychotherapy process by time-series panel analysis (TSPA) , 2009, Psychotherapy research : journal of the Society for Psychotherapy Research.

[14]  Aplicación del diseño de series temporales múltiples a un caso de intervención en dos clases de Enseñanza General Básica , 2000 .

[15]  Fred L. Collopy,et al.  Error Measures for Generalizing About Forecasting Methods: Empirical Comparisons , 1992 .

[16]  Jeff Tayman,et al.  On the validity of MAPE as a measure of population forecast accuracy , 1999 .

[17]  Jeff Tayman,et al.  A note on the measurement of accuracy for subnational demographic estimates , 2000, Demography.

[18]  Philip D. Wasserman,et al.  Neural computing - theory and practice , 1989 .

[19]  B. Ripley,et al.  Robust Statistics , 2018, Encyclopedia of Mathematical Geosciences.

[20]  Jeff Swanson,et al.  In search of the ideal measure of accuracy for subnational demographic forecasts , 1999 .

[21]  J. Guydish,et al.  Investigating the Effects of San Francisco's Treatment on Demand Initiative on a Publicly-Funded Substance Abuse Treatment System: A Time Series Analysis , 2009, Journal of psychoactive drugs.

[22]  R. R. Hamm,et al.  An assessment of the accuracy of a regional economic-demographic projection model , 1984, Demography.

[23]  Robert Fildes,et al.  The evaluation of extrapolative forecasting methods , 1992 .

[24]  Thomas Kolarik,et al.  Time series forecasting using neural networks , 1994, APL '94.

[25]  Ken-ichi Funahashi,et al.  On the approximate realization of continuous mappings by neural networks , 1989, Neural Networks.

[26]  Frederick R. Forst,et al.  On robust estimation of the location parameter , 1980 .

[27]  Steven C. Wheelwright,et al.  Forecasting methods and applications. , 1979 .

[28]  S K Smith,et al.  Evaluating the forecast accuracy and bias of alternative population projections for states. , 1992, International journal of forecasting.

[29]  Spyros Makridakis,et al.  Accuracy measures: theoretical and practical concerns☆ , 1993 .

[30]  R. Wilcox.  Introduction to Robust Estimation and Hypothesis Testing , 1997 .

[31]  C. Lewis.  Industrial and business forecasting methods : a practical guide to exponential smoothing and curve fitting , 1982 .

[32]  D. Cox,et al.  An Analysis of Transformations , 1964 .

[33]  Mitchell B. Chamlin,et al.  A time-series analysis of the impact of heavy drinking on homicide and suicide mortality in Russia, 1956-2002. , 2006, Addiction.

[34]  Adamantios Diamantopoulos,et al.  Towards a taxonomy of forecast error measures a factor‐comparative investigation of forecast error dimensions , 1994 .

[35]  D A Ahlburg,et al.  Simple versus complex models: evaluation, accuracy, and combining. , 1995, Mathematical population studies.

[36]  John W. Tukey,et al.  Exploratory Data Analysis. , 1979 .

[37]  John Snyder,et al.  Using a neural network to forecast visitor behavior , 1996 .

[38]  K. A. Loparo,et al.  Nonlinear dynamical analysis of the neonatal EEG time series: The relationship between sleep state and complexity , 2008, Clinical Neurophysiology.

[39]  Andrew M. Isserman,et al.  The Accuracy of Population Projections for Subcounty Areas , 1977 .

[40]  Kurt Hornik,et al.  Multilayer feedforward networks are universal approximators , 1989, Neural Networks.

[41]  L. Ren,et al.  Applicability of the Revised Mean Absolute Percentage Errors (MAPE) Approach to Some Popular Normal and Non-normal Independent Time Series , 2009 .

[42]  Paul A. Fishwick,et al.  Time series forecasting using neural networks vs. Box- Jenkins methodology , 1991, Simul..