Modelling the Frequency of Depression using Holt-Winters Exponential Smoothing Method

Abstract

Introduction: Depression is a worldwide concern causing lots of adverse social consequences. Aim: To assess the stochastic processes of depression frequency through statistical modelling. Materials and Methods: The data contained the processes of depression frequency in Hamadan from April 2008 to March 2016. A total of 3252 patients, according to psychiatrist diagnosed based on DSM-IV criteria, were accounted. To model and forecast the depression incidence frequency, the time series analysis method, Holt-Winter exponential smoothing, was applied. Results: After a considerable decrease in the last months of 2010 which was followed by an increase in early 2011, a constant series of frequencies happened until 2015. The first months of 2015 are the beginning of the increase in the development of depression. Conclusion: Present findings demonstrated an increase in the incidence of depression during 2016-2018. The Holt-winter exponential smoothing method can be used as an appropriate statistical tool to analyse and forecast such data.

Keywords

• depression
• incidence
• time series
• stochastic process