A study on analysis of packet amount of Naver's mobile portal

Abstract

The purpose of this paper is to build a model of packet amount of Naver mobile portal. We collected 2004 cases by measuring the sixth per access from September, 2012 to October, 2015. We use regression model with autoregressive errors, in which predictors incorporated into the model were replication, date, time, week, month. It has been found the model which errors follow AR(36), based on AIC and adjusted R 2 . We found some characteristics from our model as follows. In addition to model building, we also have discussed some meaningful features yielded from the selected model in this paper. Considering the importance of this topic, continuous researches are needed.