A Holistic Federated Learning Approach Employing Neural Networks for Climate Control within Smart Buildings

Comfort is a universally understood concept of physical and mental well-being. Maximizing comfort is of paramount importance to everyone be it at home or in a workspace. Smart Buildings offers the technology to facilitate a greater enhancement of comfort in every occupied space. This research delves into the realm of intelligent climate control within smart buildings by leveraging new machine-learning techniques. This research proposes a holistic approach to climate control by incorporating federated learning, to learn the user’s preferences across many buildings. The federated learning framework ensures that data privacy is maintained at all times. A global neural network regression model has been created that has trained on a wide variety of input data. The study contributes to the evolving field of smart building technologies, offering a sustainable and efficient solution for enhanced living and climate control.