Open AccessMemprediksi Harga Komoditas Cabe Menggunakan Metode Backpropagation di Wilayah Kota Payakumbuh
Author(s) -
Allans Prima Aulia,
- Yuhandri,
Fhajri Arye Gemilang
Publication year - 2021
Publication title -
jurnal komtekinfo
Language(s) - English
Resource type - Journals
ISSN - 2502-8758
DOI - 10.35134/komtekinfo.v8i1.96
Subject(s) - backpropagation , value (mathematics) , indonesian , statistics , commodity , agricultural science , business , mathematics , artificial neural network , computer science , environmental science , artificial intelligence , linguistics , philosophy , finance
Chili is one of the spices needed by the majority of Indonesian people. These high needs have an impact on the price of this agricultural commodity which has become very fluctuated. This study, uses the backpropagation method to predict chilli prices in Payakumbuh City, with data sourced from the Badan Pusat Statistik Kota Payakumbuh. The data format are weekly chilli price data for the period 2014 to 2019. Data variables are arranged into time series forms with 4 input values from each week per year, and 1 target value. From the test results obtained the MSE value (Mean Squared Error) of 0.00118 with prediction accuracy of 98.56%. The results of this study can prove that Artificial Neural Networks using the backpropagation method can predict commodity prices for chilli in Payakumbuh City with a good level of accuracy, so that it can be used for the following year.