The explosive growth of smart devices, network access points, and new mobile applicationrndevelopment drives users to use more and more mobile applications and, this hasrnlead to the explosive growth of mobile data trafï¬c. It has a high impact on mobile servicernproviders to manage network data trafï¬c because application usage is different from onernlocation to other with time. Understanding the application-level trafï¬c patterns from arncompletely different location angle is effective for operators and content providers to createrntechnical and business plans.rnIn this paper, we have established several typical trafï¬c patterns and predict applicationrncategory trafï¬c demand per clustered location in a mobile cellular network. Wernexplore mobile trafï¬c patterns by clustering each application category into ï¬ve clustersrnbased on trafï¬c volume and location. Then, we implement a random forest model tornpredict the trafï¬c demand of three of the most highly utilized applications per clusterrnlocation.This outcome could be useful in relevant future applications, with the prospectrnto achieve average 96% predictive accuracy per application category per cluster.rnUnderstanding popular application at the clustered locations and predicting the trafï¬crndemand of a popular application could signiï¬cantly improve user experience, averagernlatency, energy consumption, spectral efï¬ciency, back-haul trafï¬c, and network capacity.rnThose outcomes are possible via designing and implementing a cache server or planningrnand optimizing the network resources based on predicted trafï¬c demand.