Comparison of forescasting methods for interval-valued time series

Abstract

This study examines the interval-valued time series which is an ongoingness issue in time series. The study aims to obtain new time series forecasting methods using different combination of several analysis methods and modeling techniques and to determine the methods and models that provide the optimal accuracy by comparing the forecasting accuracy of the proposed methods. Different approaches and appropriate modeling techniques are used for analyzing interval-valued time series. The data of experimental study is obtained by analyzing the interval-valued time series via forecasting methods. These values represent variables in the variance analysis of our study. The obtained results are analyzed by using statistical analysis. Kruskal-Wallis H test and Mann-Whitney U-test are applied to evaluate the performance. Which one of the methods provided better forecasting results and their pros and cons are examined. Four real time series are used in the implementation and the forecasting performance of the methods are compared and evaluated.