Integrated Map-Based System Using Decision Trees and Random Forests to Classify Network Service Issues in Telecommunications Infrastructure

Telecommunications is the process of sending or receiving information in various forms and has an important role in everyday life. PT. XYZ, as one of the well-known telecommunications companies in Indonesia, provides 4G/LTE telecommunications services to all people in Indonesia. However, PT. XYZ has problems monitoring the quality of the network transmitted by its towers and often must wait for complaints from users. To overcome this problem, a map-based, and machine learning integration system was developed. This system uses map and machine learning predictive capabilities to provide BTS information and predict the locations of network problems. Predictions use network data as input for Decision Tree and Random Forest models, while geographic variables are used for map-based systems. Based on trial results, map and machine learning systems show potential in predicting user location with network issues. This system uses Decision Tree and Random Forest algorithms after optimization, achieving an accuracy of 99.4% and 99.5% respectively. This system aids PT. XYZ efficiently resolves network problems for its users and enhances service quality.