摘要：AbstractSeveral time series forecasting methods can be found in the literature. Most methods depend on a predictor of some kind to estimate parameters and the forecast values. However, no measures are available in the literature for the reliability of the predictor. This paper proposes that forecasts be viewed as functions in a Statistical Metric Space (SMS) or in a Statistical Semi-metric Space (SSMS) and suggests a method to estimate a reliability measure of the predictor. A statistical metric/semi-metric space is constructed from the time-series data. A method is proposed to construct the distribution function of the SMS/SSMS in a natural way that can quantify the reliability of the forecast. The method presented in this paper is easy to implement as a computer program.
关键词：KeywordsTime SeriesForecastingConcordanceStatistical Metric SpacesKendall'sτGINI's mean difference
