<![CDATA[Uncontrolled water consumption is a serious challenge, especially in small businesses like coffee shops. Excessive water use can lead to waste and financial losses. To address this issue, IoT (Internet of Things) technology and data analysis are applied to monitor and predict water consumption. In this study, predictive models such as Random Forest, XGBoost, and LSTM are used to analyze water consumption data. The results show that Random Forest has the best performance with the lowest prediction error and the highest R-squared value, indicating this model’s capability to explain nearly all the variance in water consumption data. Random Forest and XGBoost perform well as they can handle data with non linear features and complex interactions, while LSTM's lower performance is likely due to limited data and suboptimal hyperparameter tuning. The implementation of green accounting in this system enables effective tracking of water consumption costs. Suggested improvements include further exploration of LSTM hyperparameters, the use of ensemble techniques, and cost sensitivity analysis for water-saving policy decisions. This model is expected to provide an effective water saving solution for coffee shop owners.]]>
Copyrights © 2025
