Evidence record · UDINUS · SDGs
- Unit
- Rektorat UDINUS
- Period
- 2024
- Goals
- Record
- de3839f0-8320-4aed-9590-7aab46adf6a6
- Published
- 2026-10-30
Universitas Dian Nuswantoro conducted research on demand prediction for food and beverage small and medium-sized enterprises using sales data and weather data. The activity involved Alzami, Farrikh; Salam, Abu; Rizqa, Ifan; Irawan, Candra; Andono, Pulung Nurtantio; Aqmala, Diana; and Sartika, Mila as participants. The research addressed fluctuating demand faced by SMEs and the importance of accurate demand prediction for production, inventory, and resource allocation.
The activity used sales data together with average temperature and average humidity data from January to June 2023. It applied two demand-prediction approaches: the Seasonal Autoregressive Integrated Moving Average with Exogenous Variables (SARIMAX) model and the Vector Autoregression (VAR) model. SARIMAX was evaluated using parameters d=1, D=1, p=2, q=3, P=2, Q=2, and s=7. The source states that SARIMAX with these optimum parameters outperformed optimized VAR in predicting demand for food and beverage SMEs.
The reported SARIMAX results included an AIC of 1070.11, MSE of 80.393, MAE of 7.513, and RMSE of 8.966. The source also states that SARIMAX reduced MSE by 86.35% compared to VAR. The research highlights the relevance of external factors such as weather data in demand prediction and identifies understanding demand patterns as important for informed decisions and efficient operational optimization. Through this activity, Universitas Dian Nuswantoro carried out institutional research focused on analytical methods for SME demand prediction.