Abstract: This research explores the integration of transfer learning within optimization algorithms Genetic Algorithm (GA), Particle Swarm Optimization (PSO), and Non-dominated Sorting Genetic Algorithm II (NSGA-II) over Fuzzy Systems (FIS) for enhancing energy efficiency in assisted living environments. We optimize FIS models by testing various transfer learning combinations: GA to PSO, GA to NSGA-II, PSO to NSGA-II, and GA to NGSA-II. Results show that PSO to NSGA-II delivers the best performance. GA to NSGA-II also showed notable improvement, benefiting from NSGA-II's efficient Pareto front exploration following GA's broad search capabilities. GA to PSO demonstrated slight improvement over GA alone, but PSO after GA performed worse due to premature convergence and reduced genetic diversity. In contrast, GA to NSGA-II retained better solution diversity, improving multi-objective optimization outcomes. These findings highlight the potential of transfer learning to enhance energy optimization in complex assisted living systems and provide deeper insights into its role in improving energy efficiency through strategic algorithmic pairing.

Keywords: Transfer learning fuzzy inference systems, Energy optimization, Genetic algorithms, PSO, Multi-objective optimization, Pareto front.

_________________________________________________________________________________________