PERBANDINGAN PERFORMA ARIMA, SARIMA, DAN KNN REGRESSION DALAM MEMPREDIKSI PENJUALAN PART NUMBER DI INDUSTRI MANUFAKTUR OTOMOTIF (STUDI KASUS : PT TRI CENTRUM FORTUNA)

NURUL ANDINI, . (2025) PERBANDINGAN PERFORMA ARIMA, SARIMA, DAN KNN REGRESSION DALAM MEMPREDIKSI PENJUALAN PART NUMBER DI INDUSTRI MANUFAKTUR OTOMOTIF (STUDI KASUS : PT TRI CENTRUM FORTUNA). Sarjana thesis, UNIVERSITAS NEGERI JAKARTA.

	Text COVER.pdf Download (1MB)
	Text BAB I.pdf Download (509kB)
	Text BAB II.pdf Restricted to Registered users only Download (708kB) | Request a copy
	Text BAB III.pdf Restricted to Registered users only Download (607kB) | Request a copy
	Text BAB IV.pdf Restricted to Registered users only Download (1MB) | Request a copy
	Text BAB V.pdf Restricted to Registered users only Download (442kB) | Request a copy
	Text DAFTAR PUSTAKA.pdf Download (419kB)
	Text LAMPIRAN.pdf Restricted to Registered users only Download (2MB) | Request a copy

Abstract

Perencanaan produksi dan manajemen persediaan yang efisien menjadi tantangan utama dalam industri manufaktur otomotif, terutama dalam menjaga ketersediaan part number sesuai kebutuhan produksi. Penelitian ini bertujuan untuk membandingkan performa tiga model prediksi, yaitu ARIMA, SARIMA, dan K-Nearest Neighbor (KNN) Regression dalam memproyeksikan penjualan part number di PT Tri Centrum Fortuna. Data penjualan harian periode Januari 2022 hingga November 2024 digunakan sebagai dasar dalam membangun model time series forecasting. Proses penelitian meliputi data preparation, preprocessing, transformasi stasioneritas menggunakan uji Augmented Dickey-Fuller (ADF) dan differencing, serta pembagian data menjadi data latih dan data uji. Evaluasi performa model dilakukan dengan menggunakan empat metrik, yaitu Root Mean Square Error (RMSE), Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), dan Mean Squared Error (MSE). Hasil penelitian menunjukkan bahwa model SARIMA memiliki performa terbaik dalam mengenali pola musiman mingguan dan fluktuasi penjualan, dengan nilai kesalahan prediksi yang paling rendah pada keempat metrik evaluasi. Temuan ini mengindikasikan bahwa SARIMA merupakan model yang paling andal untuk kebutuhan peramalan penjualan part number dan dapat dijadikan dasar pengambilan keputusan dalam perencanaan stok yang lebih akurat dan efisien.***** Efficient production planning and inventory management remain major challenges in the automotive manufacturing industry, particularly in ensuring the availability of part numbers aligned with production demands. This study aims to compare the performance of three forecasting models—ARIMA, SARIMA, and K-Nearest Neighbor (KNN) Regression—in predicting part number sales at PT Tri Centrum Fortuna. Daily sales data from January 2022 to November 2024 were utilized as the foundation for time series forecasting models. The research process included data preparation, preprocessing, stationarity transformation using the Augmented Dickey-Fuller (ADF) test and differencing, and data splitting into training and testing sets. The models' performance was evaluated using four metrics: Root Mean Square Error (RMSE), Mean Absolute Error (MAE), Mean Absolute Percentage Error (MAPE), and Mean Squared Error (MSE). The results indicate that the SARIMA model achieved the best performance in capturing weekly seasonal patterns and sales fluctuations, with the lowest error values across all evaluation metrics. These findings suggest that SARIMA is the most reliable model for forecasting part number sales and can serve as a robust foundation for more accurate and efficient inventory planning decisions.

Item Type:	Thesis (Sarjana)
Additional Information:	1). Ali Idrus, M.Kom. ; 2). Lipur Sugiyanta, Ph.D.
Subjects:	Teknologi dan Ilmu Terapan > Teknik Komputer
Divisions:	FT > S1 Sistem dan Teknologi Informasi
Depositing User:	Nurul Andini .
Date Deposited:	24 Jul 2025 01:42
Last Modified:	24 Jul 2025 01:42
URI:	http://repository.unj.ac.id/id/eprint/56686

Actions (login required)

View Item