In multi-zone heating, ventilation, and air conditioning (HVAC) systems, achieving energy efficiency while maintaining occupant comfort is challenging due to inter-zone thermal interactions, heterogeneous user preferences, and the complexity of distributed temperature control. Traditional control strategies often struggle to balance comfort and energy objectives across varying scenarios and may exhibit unstable performance. This paper proposes a Multi-Preference Reinforcement Learning control method, which integrates a range of comfort–energy preferences directly into the training of a reinforcement learning agent. This approach enables adaptive policy adjustments based on user preferences, coordinating local temperature control while optimizing global energy consumption. Evaluations on a six-zone building model using EnergyPlus simulations shows that, compared with baseline methods, the proposed approach achieves up to 38.46% energy savings, maintains temperature deviations within $$0.02\,^\circ \textrm{C}$$ , and preserves occupant comfort for 96% of the time. These results demonstrate the effectiveness and robustness of this preference-aware reinforcement learning framework for intelligent, user-adaptive HVAC control.

This article proposes a deep learning solution to the online portfolio selection problem based on learning a latent structure directly from a price time series. It introduces a novel wealth flow matrix for representing a latent structure that has special regular conditions to encode the knowledge about the relative strengths of assets in portfolios. Therefore, a wealth flow model (WFM) is proposed to learn wealth flow matrices and maximize portfolio wealth simultaneously. Compared with existing approaches, our work has several distinctive benefits: (1) the learning of wealth flow matrices makes our model more generalizable than models that only predict wealth proportion vectors, and (2) the exploitation of wealth flow matrices and the exploration of wealth growth are integrated into our deep reinforcement algorithm for the WFM. These benefits, in combination, lead to a highly-effective approach for generating reasonable investment behavior, including short-term trend following, the following of a few losers, no self-investment, and sparse portfolios. Extensive experiments on five benchmark datasets from real-world stock markets confirm the theoretical advantage of the WFM, which achieves the Pareto improvements in terms of multiple performance indicators and the steady growth of wealth over the state-of-the-art algorithms.