| Title |
Nonlinear Relationship Analysis of Factors Affecting Apartment Common Maintenance Costs |
| Authors |
최규진(Choi, Kyujin) ; 김서영(Kim, Seoyoung) ; 이다니엘(Lee, Daniel) ; 오경호(Oh, Kyungho) ; 손동욱(Sohn, Dong-Wook) |
| DOI |
https://doi.org/10.5659/JAIK.2026.42.1.51 |
| Keywords |
Apartment Complex; Maintenance cost; Economies of Scale; Complex size; Number of units; Building Age; Machine learning; Nonlinear relationship |
| Abstract |
In South Korea, apartment complexes dominate the housing landscape, and rising common maintenance fees have emerged as a significant
social concern. This study examines the factors influencing common maintenance costs in apartment complexes and investigates their nonlinear
patterns using XGBoost and SHAP. The analysis highlights several findings. Building age shows a threshold effect, with costs increasing
sharply after 30 years. Complexes with more than 450 units demonstrate lower costs, while the number of buildings has little impact. Taller
buildings, particularly those exceeding 27 floors, are linked to higher expenses. In terms of management, contracted services and individual
heating systems are associated with lower costs, whereas sale-based ownership tends to raise them. Layout also plays a significant role:
staircase-type corridors are associated with higher costs, as are complexes with parking capacity above 1.2 vehicles per unit or CCTV
coverage above 0.2 cameras per unit. The absence of community facilities further contributes to increased costs. By moving beyond linear
assumptions, this study identifies thresholds and inflection points in critical variables, thereby offering practical insights for housing policy,
remodeling priorities, management efficiency, and cost-conscious apartment design. |