Modelagem Arima: uma aplicação na análise e previsão do volume de clientes do serviço telefônico pós-pago profissionais das TDM, sarl, na província de Maputo

Abstract

The aim of this study was to identify a forecast model for the volume of professional postpaid telephone service customers of the telecommunications company in mozambique, sarl, as an auxiliary tool for planning their actions. As the volume of customers is one of the factors that dictate the development of the company, it is of interest to know the future behavior of the volume of customers in particular. For this, the statistical methodology for time series was used, more specifically the box-jenkins method. The processed data were provided by tdm, sarl, commercial department, and refer to the monthly values ​​of the number of customers of professional postpaid telephone services, in the province of Maputo, during the period 2001 to August 2008. To carry out the analysis , the computer applications eviews and spss were used. Several prediction models were evaluated, resulting in a prediction model. In order to achieve the objective, the analysis of the time series was initially made using the arima methodology - box-jenkins models, arriving at the definition of mathematical models that can be used for forecasting. According to the results of the analyses, the series of customer volume can be represented by the arima model (7,1.0); Forecasts for the period from 05/2008 to 12/2008, using the arima model (7.1.0) do not differ significantly from the values ​​observed for the volume of customers