Predictive Machine Learning and Multi-Criteria Optimization of Thermal Insulation Materials for Building Energy Efficiency

Building insulation plays a pivotal role in achieving sustainable and energy-efficient infrastructure. The selection of optimal insulation materials involves multiple conflicting criteria, making it a complex decision-making problem. Traditional selection methods often lack robustness and fail to incorporate adaptive learning mechanisms, leading to inconsistencies in material rankings. This study introduces an integrated Hybrid AI-MCDM Framework, combining Analytic Hierarchy Process (AHP), Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS), Fuzzy Logic, and Machine Learning (ML-MCDM) to enhance decision accuracy and minimize uncertainty in insulation material selection. The proposed methodology involves pairwise comparisons through AHP, distance-based ranking using TOPSIS, uncertainty modeling via Fuzzy Logic, and AI-driven adaptive modeling to optimize material selection. Sensitivity analysis and robustness evaluation further validate the stability of ranking results. The findings reveal that Mineral Wool (TOPSIS: 0.88, Fuzzy: 0.85) and Aerogel (TOPSIS: 0.83, Fuzzy: 0.92) are the top-performing insulation materials, while Polystyrene (XPS) and Polyurethane exhibit the lowest suitability due to their poor fire resistance and high embodied energy. The AI-driven approach significantly enhances ranking consistency, making it a valuable tool for sustainable building insulation selection. The study underscores the advantages of integrating expert-driven MCDM techniques with AI-based learning models, ensuring a more data-driven, reliable, and adaptive decision-support system for material selection in energy-efficient buildings.