Article Per Year (5 Year)
p-Index From 2021 - 2026
0.23
P-Index
This Author published in this journals
All Journal Journal of Social Research
Muhammad Reyhan Dwirama
Universitas Gunadarma
Humanities Computer Science & IT Economics, Econometrics & Finance Education Social Sciences

Published : 1 Documents Claim Missing Document
Claim Missing Document
Check
Articles

Found 1 Documents
Search

 1 
A Comparative Analysis of the Accuracy of the ARIMA, LSTM, and Holt-Winters Methods in Forecasting SSD Prices Based on Historical Marketplace Data Muhammad Reyhan Dwirama
Journal of Social Research Vol. 5 No. 9 (2026): Journal of Social Research
Publisher : International Journal Labs (AHU-0028405-AH.01.14 Tahun 2022)

Show Abstract | Download Original | Original Source | Check in Google Scholar | DOI: 10.55324/josr.v5i9.3354

Abstract

The fluctuation of Solid-State Drive (SSD) prices, influenced by market demand, NAND Flash availability, technological developments, and the increasing need for storage in Artificial Intelligence (AI) and data centers, has become a challenge in determining purchasing timing and sales strategies. Accurate price prediction is required to assist consumers and business actors in making better decisions. This study aims to compare the performance of three time series forecasting models, namely ARIMA, Long Short-Term Memory (LSTM), and Holt-Winters Exponential Smoothing, in predicting the daily prices of 512 GB NVMe SSDs. The data used in this study consist of historical SSD price data collected from Shopee, Tokopedia, and eBay during the period of December 2025 to May 2026, which were converted into USD for consistency. The research process includes data collection, preprocessing, implementation of the three models, and performance evaluation using RMSE, MAE, and MAPE metrics. The results show that the LSTM model achieved the best performance with a MAPE value of 8.10%, compared to ARIMA at 10.88% and Holt-Winters at 11.87%. These findings indicate that LSTM is more effective in capturing complex and fluctuating data patterns, making it more optimal for SSD price prediction.
Co-Authors