PERAMALAN METODE TIME SERIES TERHADAP PRODUKSI KAKAO DI KABUPATEN BATANG

Authors

  • FADILLAH SANTIKA Teknik Industri, Institut Teknologi Nasional Bandung
  • Dwi Kurniawan Teknik Industri, Institut Teknologi Nasional Bandung

Keywords:

Kakao, Peramalan, Time Series, Moving Average, Regresi Linear, Exponential Smoothing

Abstract

ABSTRAK
Kakao merupakan salah satu komoditas perkebunan yang memiliki peran penting bagi perkembangan ekonomi di Kabupaten Batang. Produksi Kakao di Kabupaten Batang mengalami tren yang fluktuatif, dikarenakan adanya faktor seperti kebijakan dan strategi produksi dalam bersaingnya produksi kakao di Kabupaten Batang pada industri kakao. Peramalan hasil produksi diperlukan untuk mengetahui perkembangan hasil produksi dimasa depan untuk itu dengan adanya metode peramalan time series dari data masa lampau. Hasil peramalan produksi kakao pada tahun 2021 di ketiga metode didapatkan nilai pada metode MA sebesar 220.394,5 kwintal dengan nilai error sebesar MAD = 115,358, MSE = 15.298.686.354, dan MAPE = 43,05%. Nilai peramalan periode 2021 pada metode ES sebesar 261.664,5 kwintal dengan nilai error sebesar MAD = 97.898,6 , MSE = 12.296.282.040, dan MAPE = 36,91%. Nilai peramalan periode 2021 pada metode LR sebesar 165.485,7 kwintal dengan nilai error sebesar MAD = 42.329,8, MSE = 3.25.722.398, dan MAPE = 11,56%. Linear Regression merupakan metode peramalan yang terpilih dikarenakan tingkat keakurasian  peramalan paling baik yang disebabkan nilai error didapatkan paling minimal dengan nilai MAD = 42.329,8, MSE = 3.25.722.398, dan MAPE = 11,56%.

 

ABSTRACT
Cocoa is one of the plantation commodities that has an important role for economic development in Batang Regency. Cocoa production in Batang Regency experiences a fluctuating trend, due to factors such as policies and production strategies in competing cocoa production in Batang Regency in the cocoa industry. Forecasting production results is needed to determine the development of production results in the future for that with the time series forecasting method from past data. The results of forecasting cocoa production in 2021 in the three methods obtained a value in the MA method of 220,394.5 quintals with an error value of MAD = 115,358, MSE = 15,298,686.354, and MAPE = 43,05%. The forecast value for the 2021 period in the ES method is 261,664,5 quintals with an error value of MAD = 97,898.6 , MSE = 12,296,282,040, and MAPE = 36.91%. The forecast value for the 2021 period in the LR method is 165,485.7 quintals with an error value of MAD = 42,329.8, MSE = 3.25,722,398, and MAPE = 11.56%. Linear Regression is the chosen forecasting method because the level of forecasting accuracy is the best due to the minimum error value obtained with MAD = 42,329.8, MSE = 3.25,722,398, and MAPE = 11.56%.

Published

2022-06-21 — Updated on 2022-07-11

Versions