PERAMALAN PENJUALAN AKI DENGAN MODEL ARIMA SEBAGAI DASAR REKOMENDASI SAFETY STOCK PADA PT RESTU AKI DENGAN FRAMEWORK CRISP-DM

Herdiyanto, Rafi Argya Darma (2026) PERAMALAN PENJUALAN AKI DENGAN MODEL ARIMA SEBAGAI DASAR REKOMENDASI SAFETY STOCK PADA PT RESTU AKI DENGAN FRAMEWORK CRISP-DM. Undergraduate thesis, Universitas Pembangunan Nasional Veteran Jawa Timur.

	Text (Cover) 22082010116_Cover.pdf Download (2MB)
	Text (Bab I) 22082010116_Bab I.pdf Download (505kB)
	Text (Bab II) 22082010116_Bab II.pdf Restricted to Repository staff only until 28 October 2029. Download (977kB) \| Request a copy
	Text (Bab III) 22082010116_Bab III.pdf Restricted to Repository staff only until 28 October 2029. Download (1MB) \| Request a copy
	Text (Bab IV) 22082010116_Bab IV.pdf Restricted to Repository staff only until 28 October 2029. Download (2MB) \| Request a copy
	Text (Bab V) 22082010116_Bab V.pdf Download (490kB)
	Text (Daftar Pustaka) 22082010116_DaftarPustaka.pdf Download (431kB)
	Text (Lampiran) 22082010116_Lampiran.pdf Restricted to Repository staff only until 28 October 2029. Download (795kB) \| Request a copy

Abstract

Demand uncertainty is a challenge in inventory management at retail businesses, including Toko Aki Restu. Fluctuations in demand for each battery brand can create discrepancies between stock availability and customer needs, increasing the risk of overstock and stockout. This study aims to forecast monthly battery sales by brand using the Autoregressive Integrated Moving Average (ARIMA) method and use the results to determine Safety Stock and Reorder Point (ROP). The research follows the Cross-Industry Standard Process for Data Mining (CRISP-DM) framework, comprising Business Understanding, Data Understanding, Data Preparation, Modeling, Evaluation, and Deployment. Historical sales transaction data from July 1, 2016, to July 31, 2026, were aggregated by month and battery brand. The dataset was divided into training data (80%, 873 rows) and testing data (20%, 216 rows). Modeling involved stationarity testing using the Augmented Dickey-Fuller (ADF) test, parameter identification using the Autocorrelation Function (ACF) and Partial Autocorrelation Function (PACF), and Grid Search to identify optimal ARIMA parameters within a range of 0 to 15. The results produced different best-fit ARIMA models for nine battery brands: AMARON ARIMA(4,0,4), BOSCH ARIMA(0,0,3), BOSCHAGM ARIMA(2,0,2), DELKOR ARIMA(3,0,3), FURUKAWA BATTERY ARIMA(2,0,1), GS ASTRA ARIMA(0,0,2), GS ASTRA MF ARIMA(2,0,4), RCA BATTERY ARIMA(3,0,3), and SUPREME ARIMA(1,0,9). Testing results showed Mean Absolute Error (MAE) values of 1.90–5.50 units, Root Mean Squared Error (RMSE) values of 2.30–6.74 units, and Mean Absolute Percentage Error (MAPE) values of 13.26–41.24%. Forecasts for August 2026 to July 2027 were used to calculate Safety Stock and ROP based on a seven-day procurement lead time and a 95% service level. The resulting Safety Stock ranged from 2 to 6 units, while ROP ranged from 3 to 16 units across brands. Finally, the forecasts, ARIMA models, evaluation metrics, inventory parameters, and sales details by battery type were integrated into a web-based system to support inventory analysis and decision-making at Toko Aki Restu.

Item Type:

Thesis (Undergraduate)

Contributors:

Contribution	Contributors	NIDN/NIDK	Email
Thesis advisor	Wati, Seftin Fitri Ana	NIDN0020039104	seftin.fitri.si@upnjatim.ac.id
Thesis advisor	Hafidz, Mohammad Al	NIDN0722099104	hafidz.si@upnjatim.ac.id

Subjects:

T Technology > T Technology (General) > T58.6-58.62 Management Information Systems

Divisions:

Faculty of Computer Science > Departemen of Information Systems

Depositing User:

Rafi Argya Darma Herdiyanto

Date Deposited:

28 Oct 2026 03:23

Last Modified:

28 Oct 2026 03:23

URI:

https://repository.upnjatim.ac.id/id/eprint/61089