Knowledge-Based Systems, Год журнала: 2024, Номер 304, С. 112400 - 112400
Опубликована: Сен. 6, 2024
Язык: Английский
Knowledge-Based Systems, Год журнала: 2024, Номер 304, С. 112400 - 112400
Опубликована: Сен. 6, 2024
Язык: Английский
Sustainable Cities and Society, Год журнала: 2023, Номер 99, С. 104946 - 104946
Опубликована: Сен. 18, 2023
Язык: Английский
Процитировано
60Energy Conversion and Management, Год журнала: 2024, Номер 306, С. 118207 - 118207
Опубликована: Март 16, 2024
Язык: Английский
Процитировано
55Journal of Building Engineering, Год журнала: 2023, Номер 73, С. 106735 - 106735
Опубликована: Май 9, 2023
Язык: Английский
Процитировано
54Sustainable Cities and Society, Год журнала: 2023, Номер 101, С. 105107 - 105107
Опубликована: Дек. 13, 2023
Язык: Английский
Процитировано
46Applied Soft Computing, Год журнала: 2024, Номер 154, С. 111318 - 111318
Опубликована: Фев. 2, 2024
Язык: Английский
Процитировано
18Sustainable Cities and Society, Год журнала: 2024, Номер 103, С. 105256 - 105256
Опубликована: Фев. 7, 2024
Язык: Английский
Процитировано
17Journal of Environmental Management, Год журнала: 2024, Номер 364, С. 121264 - 121264
Опубликована: Июнь 12, 2024
The considerable amount of energy utilized by buildings has led to various environmental challenges that adversely impact human existence. Predicting buildings' usage is commonly acknowledged as encouraging efficiency and enabling well-informed decision-making, ultimately leading decreased consumption. Implementing eco-friendly architectural designs paramount in mitigating consumption, particularly recently constructed structures. This study utilizes clustering analysis on the original dataset capture complex consumption patterns over periods. yields two distinct subsets represent low high an additional subset exclusively encompasses weekends, attributed specific behavior occupants. Ensemble models have become increasingly popular due advancements machine learning techniques. research three discrete algorithms, namely Artificial Neural Network (ANN), K-nearest neighbors (KNN), Decision Trees (DT). In addition, application employs more algorithms bagging boosting: Random Forest (RF), Extreme Gradient Boosting (XGB), (GBT). To augment accuracy predictions, a stacking ensemble methodology employed, wherein forecasts generated many are combined. Given obtained outcomes, thorough examination undertaken, encompassing techniques stacking, bagging, boosting, conduct comprehensive comparative study. It pertinent highlight technique consistently exhibits superior performance relative alternative methodologies across spectrum heterogeneous datasets. Furthermore, using genetic algorithm enables optimization combination base learners, resulting notable enhancement prediction accuracy. After implementing this technique, GA-Stacking demonstrated remarkable Mean Absolute Percentage Error (MAPE) scores. improvement observed was substantial, surpassing 90 percent for all subset-1, subset-2, subset-3, achieved R
Язык: Английский
Процитировано
10Energy Conversion and Management, Год журнала: 2025, Номер 327, С. 119589 - 119589
Опубликована: Фев. 1, 2025
Язык: Английский
Процитировано
2Results in Engineering, Год журнала: 2025, Номер unknown, С. 104356 - 104356
Опубликована: Фев. 1, 2025
Язык: Английский
Процитировано
2Journal of Energy Storage, Год журнала: 2025, Номер 114, С. 115850 - 115850
Опубликована: Фев. 22, 2025
Язык: Английский
Процитировано
2