| dc.description.abstract |
Currently the internet continues to roll so that it begins to be felt as a basic need to obtain new and complete information. Many people ranging from children to adults are compelled to access the internet for their various needs. This is the cause of a flood of internet data sales at various prices. Many quota-based internet packages are offered by cellular operator companies. Because quota-based internet is very suitable and in demand by people who have various interests related to internet use, the demand for internet packages is increasing with different prices for each operator. And this research has a problem that is happening at the moment, namely that there has never been a prediction of package prices for each cellular operator because prices vary and change. Predictions on internet package prices are very important when accurate prediction results can help cellular companies and traders in setting prices. therefore in this study using the GRU (Gated Recurrent Unit) Deep Learning Method for Time Series Forecasting or Time Series Forecasting from changes in internet package prices. Based on research using the GRU method, the number of hidden layer neurons with the most optimal results is 256 hidden layer neurons. This is because the 256 hidden layer neurons have the lowest error rate, namely 12,247 on the train data and 11,481 on the test data.
Keywords: internet,prediction,GRU
| en_US |
