<![CDATA[The dynamics of post-pandemic tourism recovery create significant fluctuations in Bali's hospitality industry, demanding a precise capacity management method. This study aims to build a robust forecasting model to project the hotel Room Occupancy Rate (TPK) using the Seasonal Autoregressive Integrated Moving Average (SARIMA) algorithm approach. This study utilizes monthly time series datasets sourced from the Central Bureau of Statistics (BPS) during the recovery phase, namely the period January 2022 to October 2025. The research methodology applies an analytical framework that includes stationarity test, differencing process, and identification of the optimal model through Auto-ARIMA mechanism to capture complex seasonal patterns. Model performance is validated using the statistical metrics of Mean Absolute Error (MAE), Root Mean Square Error (RMSE), and Mean Absolute Percentage Error (MAPE). The experimental results show that the SARIMA(0,1,0)(1,0,0)[12] model architecture is the best model with the lowest Akaike Information Criterion (AIC) value. The performance evaluation resulted in a MAPE value of 8.68%, which indicates a "very good" level of accuracy in minimizing prediction errors. Based on the model, projections for the period November 2025 to October 2026 show a positive stability trend with an estimated average occupancy of 63.32% (range 58.01%-66.48%). This research contributes to providing tourism stakeholders with a reliable quantitative instrument for the formulation of pricing strategies and more efficient resource allocation.]]>
Copyrights © 2026
