Hybrid LSTM–Transformer architecture for predictive indoor operative temperature modeling in sentry buildings

Key Points

• The architecture predicts indoor operative temperatures effectively, improving over traditional models.
• Key evidence shows an enhanced accuracy of 15% compared to baseline models in temperature forecasting.
• Analysis of predictive modeling using neural networks combines LSTM and transformer elements for optimal results.
• This approach supports better energy management strategies in sentry buildings, indicating potential for wider applications.