Prosiding Snastikom
Vol. 1 No. 1 (2021): Prosiding Snastikom 2021

Implementasi Model Recurrent Neural Network Dalam Melakukan Prediksi Harga Kartu Perdana Internet Dengan Menggunakan Algoritma Long Short Term Memory

Nafila Hayati Lubis (Universitas Harapan Medan)
Yessi Fitria Annisah Lubis (Universitas Harapan Medan)



Article Info

Publish Date
30 Oct 2021

Abstract

The development of the internet starter pack business has finally shown a symptom, namely the increasing number and variety of internet starter packs made by provider companies with various types, features and facilities that are more complete and bonuses given. With the Recurrent Neural Network method, it is an artificial neural network architecture that has been proven to perform well because the processing is called repeatedly to process sequential data input. Using the Long Short Term Memory algorithm to predict the price of internet starter packs that will be generated through raw data that is used as input. The results of the study predict the price of internet starter packs with 70% training data (529) and 30% test data (226) with a total of 755 data from August to December, with forecasting time series epochs 100, hidden layer 32, batch size 64 overall predictions the price of an internet starter pack is 52,500 to 67,000 using the RNN and LSTM methods

Copyrights © 2021






Journal Info

Abbrev

SNASTIKOM2020

Publisher

Subject

Computer Science & IT Library & Information Science

Description

seminar Nasional Teknologi Informasi dan Komunikasi yang diselangarakan oleh Program Studi Teknik Informatika Universitas Harapan Medan merupakan agenda kegiatan tahunan sebagai sarana pengembangan ilmu pengetahuan dibidang teknologi informasi dan komunikasi untuk mahasiswa, akademisi maupun ...