| ³í¹®¸í |
°Ç¹° ½Ã½ºÅÛ¿¡ °üÇÑ ¿Â¶óÀÎ ¸ðµ¨°ú ¿ÀÇÁ¶óÀÎ ¸ðµ¨ÀÇ ºñ±³ / Online vs. offline machine learning models for building systems / Ãß°è-05. °ÇÃàȯ°æ¹×¼³ºñ |
| ÀúÀÚ¸í |
ÃßÇѰæ(Chu, Han-Gyeong) ; ¼¿øÁØ(Suh, Won-Jun) ; ½ÅÇѼÖ(Shin, Han-Sol) ; ¶ó¼±Áß(La, Seon-Jung) ; ¹Úö¼ö(Park, Cheol-Soo) |
| ¼ö·Ï»çÇ× |
´ëÇѰÇÃàÇÐȸ Çмú¹ßÇ¥´ëȸ ³í¹®Áý, Vol.36 No.2 (2016-10) |
| ÆäÀÌÁö |
½ÃÀÛÆäÀÌÁö(645) ÃÑÆäÀÌÁö(2) |
| ÁÖÁ¦¾î |
±â°èÇнÀ ; ¿ÀÇÁ¶óÀÎ ¸ðµ¨ ; ¿Â¶óÀÎ ¸ðµ¨ ; Machine Learning ; Offline model ; Online model |
| ¿ä¾à2 |
In this study, the authors made a comparison study between on-line and offline models for five different building systems (a chiller, a cooling tower, a pump and an AHU) in an existing office building. Five different machine learning algorithms were applied to the aforementioned systems: Artificial Neural Network (ANN), Support Vector Machine (SVM), Gaussian Process Emulator (GPE), Random Forest (RF) and Genetic Programming (GP). The results show that the online models are not always better than the off-line models. It is highlighted in the paper that the online models perform worse than the offline models under unusual circumstances. |