YURI PINESTI TANTIA, NPM 2109100089 (2025) ANALISIS PREDIKSI HASIL PANEN KELAPA SAWIT DI PT SINAR PANDAWA MENGGUNAKAN METODE ALGORITMA C4.5 DAN REGRESI LINEAR. Skripsi thesis, Universitas Labuhanbatu.
![[img]](http://repository.ulb.ac.id/style/images/fileicons/text.png) |
Text
COVER.pdf Download (2MB) |
|
Text
BAB I.pdf Download (162kB) |
|
Text
BAB II.pdf Download (418kB) |
|
Text
BAB III.pdf Download (346kB) |
|
Text
BAB IV.pdf Download (2MB) |
|
Text
BAB V.pdf Download (150kB) |
|
Text
DAFTAR PUSTAKA.pdf Download (182kB) |
|
Text
LAMPIRAN.pdf Download (1MB) |
Abstract
Perkebunan kelapa sawit memiliki peran penting dalam perekonomian Indonesia, sehingga prediksi hasil panen yang akurat sangat dibutuhkan untuk mendukung perencanaan dan efisiensi produksi. Saat ini, PT Sinar Pandawa masih menggunakan metode manual dalam memperkirakan hasil panen, yang seringkali kurang akurat dan subjektif. Penelitian ini bertujuan untuk membandingkan dua metode prediksi, yaitu algoritma C4.5 dan regresi linear. Data yang digunakan berupa data historis hasil panen bulanan selama satu tahun, dengan variabel curah hujan, jumlah pupuk, dan hasil panen. Analisis menggunakan algoritma C4.5 menghasilkan model pohon keputusan dengan kategori (rendah, sedang, tinggi) dan nilai evaluasi RMSE sebesar 60.779 serta MAE sebesar 56.425. Sementara itu, model regresi linear menghasilkan persamaan Y=1159.98−0.0702X1+0.0104X2Y = 1159.98 - 0.0702X_1 + 0.0104X_2Y=1159.98−0.0702X1+0.0104X2, dengan nilai RMSE sebesar 3.545, MAE sebesar 1.039, dan R² sebesar 0.519. Hasil penelitian menunjukkan bahwa regresi linear lebih akurat dalam memprediksi hasil panen kelapa sawit dibandingkan algoritma C4.5, meskipun C4.5 tetap memiliki keunggulan dalam interpretasi aturan klasifikasi. Dengan demikian, regresi linear direkomendasikan sebagai metode prediksi utama, sedangkan C4.5 dapat digunakan sebagai pendukung dalam pengambilan keputusan berbasis kategori. Kata Kunci: Prediksi Hasil Panen, Kelapa Sawit, Algoritma C4.5, Regresi Linear, Data Mining ========================================================================================== Palm oil plantations play an important role in Indonesia’s economy, making accurate yield prediction essential to support production planning and efficiency. Currently, PT Sinar Pandawa still relies on manual methods to estimate yields, which are often inaccurate and subjective. This study aims to compare two prediction methods, namely the C4.5 algorithm and linear regression. The dataset used consists of monthly historical harvest data for one year, with rainfall, fertilizer usage, and yield as variables. Analysis using the C4.5 algorithm produced a decision tree model with categorical outputs (low, medium, high), resulting in an RMSE of 60.779 and an MAE of 56.425. Meanwhile, the linear regression model generated the equation Y=1159.98−0.0702X1+0.0104X2Y = 1159.98 - 0.0702X_1 + 0.0104X_2Y=1159.98−0.0702X1+0.0104X2, with an RMSE of 3.545, MAE of 1.039, and R² of 0.519. The results indicate that linear regression provides more accurate predictions of palm oil yields compared to the C4.5 algorithm, although C4.5 remains advantageous in providing easily interpretable classification rules. Therefore, linear regression is recommended as the primary prediction method, while C4.5 can be applied as a complementary tool for decision-making based on categorical outputs. Keywords: Yield Prediction, Palm Oil, C4.5 Algorithm, Linear Regression, Data Mining
| Item Type: | Thesis (Skripsi) |
|---|---|
| Uncontrolled Keywords: | Prediksi Hasil Panen, Kelapa Sawit, Algoritma C4.5, Regresi Linear, Data Mining================Yield Prediction, Palm Oil, C4.5 Algorithm, Linear Regression, Data Mining |
| Subjects: | Q Science > Q Science (General) Q Science > QA Mathematics > QA75 Electronic computers. Computer science T Technology > T Technology (General) |
| Divisions: | Fakultas Sains Dan Teknologi > Sistem Informasi |
| Depositing User: | Unnamed user with email repository@ulb.ac.id |
| Date Deposited: | 15 Sep 2025 04:32 |
| Last Modified: | 15 Sep 2025 04:32 |
| URI: | http://repository.ulb.ac.id/id/eprint/1684 |
Actions (login required)
 |
View Item |
