Nisaul Husna, . (2026) PENERAPAN SUPPORT VECTOR REGRESSION DALAM MEMPREDIKSI KEBUTUHAN BAHAN BAKU PADA CV. XYZ. Skripsi thesis, Universitas Pembangunan Nasional Veteran Jakarta.
![[img]](http://repository.upnvj.ac.id/style/images/fileicons/text.png) |
Text
ABSTRAK.pdf Download (221kB) |
|
Text
AWAL.pdf Download (3MB) |
|
Text
BAB I.pdf Restricted to Repository UPNVJ Only Download (307kB) |
|
Text
BAB II.pdf Restricted to Repository UPNVJ Only Download (4MB) |
|
Text
BAB III.pdf Restricted to Repository UPNVJ Only Download (463kB) |
|
Text
BAB IV.pdf Restricted to Repository UPNVJ Only Download (761kB) |
|
Text
BAB V.pdf Download (380kB) |
|
Text
DAFTAR PUSTAKA.pdf Download (240kB) |
|
Text
LAMPIRAN.pdf Restricted to Repository UPNVJ Only Download (1MB) |
|
Text
RIWAYAT HIDUP.docx.pdf Restricted to Repository UPNVJ Only Download (222kB) |
|
Text
HASIL PLAGIARISME.pdf Restricted to Repository staff only Download (21MB) |
|
Text
ARTIKEL KI.pdf Restricted to Repository staff only Download (853kB) |
Abstract
CV. XYZ is a palm oil processing company that uses loose oil palm fruits as its primary raw material. Fluctuations in raw material usage and unintegrated operational records have caused purchasing plans to remain insufficiently supported by data. This study applies Support Vector Regression (SVR) using a multi-output regression approach to predict raw material usage for the next three days and integrate the model into a web-based application. The dataset consists of 394 daily records from June 1, 2025, to June 29, 2026. The model uses 10 input features, namely beginning inventory, incoming raw materials, production output, and raw material usage on the current day and the previous six days. Hyperparameter optimization was performed using grid search with a time-series cross-validation scheme. Based on the lowest cross-validation MAE of 12,228.9836, the best model used the Polynomial kernel with a regularization parameter C of 10, epsilon of 0.05, gamma of 0.01, degree of 3, and coef0 of 1.0. The model achieved MAPE values of 6.6396% for the “H+1” target, 15.3602% for the “H+2” target, and 24.2327% for the “H+3” target. The model was integrated using React.js, FastAPI, and PostgreSQL. All 34 black-box testing scenarios passed, while UAT achieved an average score of 95%. The results indicate that the system can support inventory monitoring and raw material purchasing planning.
| Item Type: | Thesis (Skripsi) |
|---|---|
| Additional Information: | [No.Panggil: 2210511055] [Pembimbing 1: Jayanta] [Pembimbing 2: Nur Hafifah Matondang] [Penguji 1: Dr. Noor Falih] [Penguji 2: Kharisma Wiati Gusti] |
| Uncontrolled Keywords: | inventory, multi-output regression, prediction, raw materials, Support Vector Regression. |
| Subjects: | Q Science > QA Mathematics > QA75 Electronic computers. Computer science Q Science > QA Mathematics > QA76 Computer software |
| Divisions: | Fakultas Ilmu Komputer > Program Studi Informatika (S1) |
| Depositing User: | NISAUL HUSNA |
| Date Deposited: | 29 Jul 2026 13:54 |
| Last Modified: | 29 Jul 2026 13:54 |
| URI: | http://repository.upnvj.ac.id/id/eprint/52144 |
Actions (login required)
 |
View Item |