Deep learning-based method for optimizing property service experience

Modern property management systems are under increasing pressure to provide high-quality services, maximize resource utilization, and improve resident happiness. The integration of Internet of Things (IoT) devices and smart sensors allows for real-time data collection, yet traditional rule-based systems frequently fail to evaluate this data efficiently. The goal of this research is to develop an artificial intelligence (AI)-powered decision support system (DSS) based on deep learning (DL) techniques to improve service quality, predict energy and maintenance needs, and enhance the overall resident experience. A DL-based DSS integrates the Archimedes-Optimized Tabular Network (AOTNet) method to improve the property service experience through predictive analytics and intelligent resource management. Real-time information from IoT sensors, maintenance logs, and resident feedback is gathered. Data preprocessing using data cleaning and Min-max normalization scales numerical variables, ensuring uniformity and predictive accuracy. To minimize dimensionality and emphasize essential variance in energy data, Principal Component Analysis (PCA) is used. AOTNet combines the Archimedes Optimization Algorithm (AOA) for hyperparameter tuning with Tabular Network (TNet) for modeling structured tabular data. The paired t-test is used in statistical analysis to assess the significance of performance improvements. AOTNet offers values for $$\:{R}^{2}$$ (96%), MAE (0.030), RMSE (0.050), MAPE (3.2%), and MSE (214.2) while improving task completion and resident satisfaction. This research provides an intelligent and adaptable DSS for optimizing property services, which provides strong predictive insights and real-time decision support.