Prediksi Jumlah Follower Official Account Line Menggunakan Regresi dan Algoritma Genetika

Nizar Riftadhi Prabandaru (Fakultas Ilmu Komputer, Universitas Brawijaya)
Rekyan Regasari Mardhi Putri (Fakultas Ilmu Komputer, Universitas Brawijaya)
Agus Wahyu Widodo (Fakultas Ilmu Komputer, Universitas Brawijaya)

Jurnal Pengembangan Teknologi Informasi dan Ilmu Komputer
Vol 1 No 11 (2017): November 2017

This research used a case study of LINE IKI MALANG official account. The result of the interview showed that LINE IKI MALANG did not have a basis of its marketing service price change. Therefore, the author tried to provide assistance by giving a basis for any price change of the marketing service based on the predicted number of its followers on each month. The method used in this research was a regression method built with genetic algorithm. Regression was used to predict the followers, while the genetic algorithm was used to optimize the variables that influenced the predicted result. To find optimal predictive results, every optimized variable will be tested on a particular vulnerable numbers. But from the experiment, the researcher got a non-optimal result which was 7,801E-03 because of the minimum data so that the prediction was not close to the training data.