ZEMCH 2019 International Conference Proceedings April.2020 | Page 25
building characteristics are significant in both models. To be specific, number of children under 8 and
occupant density are significant in both models, yet, COP of air conditioner is significant only in the
DT model. This is because COP of air conditioner affects only the subset where children under 8 do not
reside. When the variables are under such complicated conditions, there is limitation in discovering
their relationship only by a linear model. Hence, for a deeper analysis on the determinants, it is
desirable to use both linear and nonlinear models instead of relying only on a single model.
Table 2. Comparison of determinants in multiple linear regression (MLR) and decision tree (DT) (Variables
significant in each model are expressed only)
Explanatory variables
Building Characteristics
Year of Building Permit
Surface Area
System Characteristics
Use of Aux. Heating Devices
COP of Air‐con.
Occupant Characteristics
Occupant Density
# of Under 8
# of Employed
Heating Set T
EUI for heating
MLR
DT
EUI for cooling
MLR
DT
X
X X
X X
X
X X
X X
X
X X
Acknowledgments
This research was supported by grant (19AUDP‐B079104‐06) from Architecture & Urban Development Research
Program funded by Ministry of Land, Infrastructure and Transport of Korean government.
References
1.
2.
3.
MOTIE, Energy Consumption Survey 2017, 2017.
Giorgio Besagni, Marco Borgarello, The determinants of residential energy expenditure in Italy, Energy,
Volume 165, 2018, pp. 369‐386.
Molnar Christoph, Chapter 4. Interpretable Models. Interpretable Machine Learning ‐ A Guide for Making
Black Box Models Explainable, 2019. https://christophm.github.io/interpretable‐ml‐book/ (accessed on
23.08.2019)
© 2019 by the authors. Submitted for possible open access publication under the terms
and conditions of the Creative Commons Attribution (CC BY) license
(http://creativecommons.org/licenses/by/4.0/).
Analyzing Determinants of Energy Consumption for Heating and Cooling in Apartment Units –
Comparison of Linear and Nonlinear Statistical Models
14