<![CDATA[PT. Telkomsel is one of the largest telecommunication providers in Indonesia which also has the most customers spread throughout Indonesia. Customers of PT. Telkomsel from year to year has experienced an increase and this will result in the use of an increasing number of payloads because the payload is all packages that are received and sent by mobile to a receiver (signal receiver) and if the amount of payload usage is smaller than the number of users it will occur over lagging and users will feel uncomfortable. With these problems, it requires an implementation of several methods to predict the amount of 4G payload usage so that PT. Telkomsel can find out the number of 4G payload usage in the next day or month so that it can anticipate losses or complaints from customers. Of the many prediction methods available, the authors use the backpropagation neural network method to perform a prediction process using artificial neural network architecture 4 input node neurons, 6 hidden node neurons and 1 node output neuron. By using the MAPE calculation (Mean Absolute Precentage Error) the most optimal value is 6.0154830745999%.]]>
Copyrights © 2019
