ANALISIS FORECASTING PRODUKSI JAGUNG PIPILAN KERING DI PROVINSI GORONTALO: PENDEKATAN KOMPARATIF METODE SUPPORT VECTOR REGRESSION (SVR) DAN SARIMAX

Sue, Mohamad Fikri (2026) ANALISIS FORECASTING PRODUKSI JAGUNG PIPILAN KERING DI PROVINSI GORONTALO: PENDEKATAN KOMPARATIF METODE SUPPORT VECTOR REGRESSION (SVR) DAN SARIMAX. Sarjana thesis, Universitas Muhammadiyah Gorontalo.

[img] Text (Abstrak Mohamad Fikri Sue)
Abstrak.pdf

Download (280kB)
[img] Text (Bab 1 Mohamad Fikri Sue)
BAB 1.pdf

Download (24kB)
[img] Text (Daftar Pustaka Mohamad Fikri Sue)
Daftar Pustaka.pdf

Download (117kB)
[img] Text (Skripsi Mohamad Fikri Sue)
Skripsi - Mohamad Fikri Sue.pdf
Restricted to Repository staff only

Download (8MB)

Abstract

Corn production in Gorontalo Province experiences fluctuations due to rainfall anomalies and changes in harvested land area, which affect regional food security. This study aims to compare the reliability of the Support Vector Regression (SVR) and Seasonal Autoregressive Integrated Moving Average with Exogenous Regressors (SARIMAX) methods in forecasting corn production. The study used monthly time series data from six administrative regions during the 2018–2025 period, with a total of 576 observations. The modeling was conducted using a multivariate approach involving historical production, harvested area, and satellite rainfall variables. The SVR model was optimized using the Radial Basis Function kernel and evaluated using Mean Absolute Percentage Error (MAPE). The results showed that SVR outperformed SARIMAX across all observation regions. The SVR model was able to handle extreme fluctuation patterns with error rates ranging from 5.58% to 15.01%, whereas SARIMAX produced significantly higher errors on production anomaly data. Therefore, SVR is considered more accurate in modeling the complexity of corn production and has the potential to be used as an early warning system to support regional food security policies. Keywords: Forecasting, Corn, SVR, SARIMAX, Fluctuation, Gorontalo.

Item Type: Thesis (Sarjana)
Contributors:
ContributionContributorsNIDN/NBMEmail
Thesis advisorHasyim, Wahyudin0911078705UNSPECIFIED
Thesis advisorGani, Hilmansyah1611058501UNSPECIFIED
Additional Information: Produksi jagung di Provinsi Gorontalo mengalami fluktuasi akibat anomali curah hujan dan perubahan luas lahan panen yang berdampak pada ketahanan pangan daerah. Penelitian ini bertujuan membandingkan keandalan metode Support Vector Regression (SVR) dan Seasonal Autoregressive Integrated Moving Average with Exogenous Regressors (SARIMAX) dalam meramalkan produksi jagung. Penelitian menggunakan data deret waktu bulanan dari enam wilayah administratif periode 2018–2025 dengan total 576 data. Pemodelan dilakukan secara multivariat menggunakan variabel produksi historis, luas panen, dan curah hujan satelit. Model SVR dioptimalkan menggunakan kernel Radial Basis Function dan dievaluasi menggunakan Mean Absolute Percentage Error (MAPE). Hasil penelitian menunjukkan bahwa SVR memiliki performa lebih baik dibandingkan SARIMAX pada seluruh wilayah pengamatan. Model SVR mampu menangani pola fluktuasi ekstrem dengan tingkat kesalahan antara 5,58% hingga 15,01%, sedangkan SARIMAX menghasilkan kesalahan yang jauh lebih tinggi pada data anomali produksi. Dengan demikian, SVR dinilai lebih akurat dalam memodelkan kompleksitas produksi jagung dan berpotensi digunakan sebagai sistem peringatan dini untuk mendukung kebijakan ketahanan pangan daerah.
Uncontrolled Keywords: Peramalan, Jagung, SVR, SARIMAX, Fluktuasi, Gorontalo.
Subjects: Q Science > QA Mathematics > QA76 Computer software
Divisions: Fakultas Sains dan Ilmu Komputer > Program Studi Sistem informasi
Depositing User: Mohamad Fikri Sue
Date Deposited: 08 Jul 2026 07:21
Last Modified: 08 Jul 2026 07:21
URI: http://repository.umgo.ac.id/id/eprint/6496

Actions (login required)

View Item View Item