建筑与文化Issue(2)：239-241,3.DOI:10.19875/j.cnki.jzywh.2025.02.074
基于增量决策树的可解释个体热舒适模型构建方法研究
Research on the Construction Method of Interpretable Individual Thermal Comfort Model Based on Incremental Decision Tree
摘要
Abstract
Indoor thermal comfort refers to the human perception of the thermal environment,which is closely related to health,work efficiency,and safety.Therefore,it is of great significance to study thermal comfort models that can accurately predict an individual's thermal comfort perception,treating specific individuals as the analysis object.However,existing studies often require large amounts of data before constructing individual thermal comfort models,which is not practical for real-world applications.Although some research has started to use new samples and incremental learning to build individual thermal comfort models,the interpretability of these models is too low,which may pose safety risks.Therefore,this study introduces the incremental decision tree method,which utilizes sequential data generated by a single user in daily life to build and update an interpretable individual thermal comfort model for that user.Experimental verification shows that this method can establish an interpretable individual thermal comfort model.Moreover,compared to existing methods,the model construction method proposed in this paper improves thermal comfort prediction accuracy by 44%when the number of user samples is small.关键词
热舒适/增量学习/决策树/机器学习Key words
thermal comfort/incremental learning/decision tree/machine learning引用本文复制引用
李献英..基于增量决策树的可解释个体热舒适模型构建方法研究[J].建筑与文化,2025,(2):239-241,3.基金项目
陕西省教育厅科学研究计划项目资助(项目编号:23JK0572) （项目编号:23JK0572）
