NAKHARA (Journal of Environmental Design and Planning)
Publication Date
2024
Abstract
Specifying the house type of a housing development project is extremely necessary. However, the determination of a project type nowadays has become a delicate matter, requiring the expertise and knowledge of seasoned project developers. This study aimed to apply four machine learning techniques: Decision Tree, Random Forest, Gradient Boosted Tree and Ensemble Classifier, to analyze the data from 179 housing estate projects collected from market reports of real estate companies in Thailand, with a focus on selecting projects with average monthly sales rates that are higher than the average of all acquired projects. This process resulted in a reduced dataset of 59 projects, including 31 townhouses, 22 single-family houses, and six semi-detached houses. As a result, the Ensemble Classifier model has the highest accuracy of 90.91%. The factors most influential in identifying the type of project are the distances from a main road, sky train station, bus station, hospital, and department store. Single-detached house projects are suitable for locations with high potential. The ideal location should be in proximity to a main road, bus station, department store, and hospital. In addition, townhouse projects are ideal for medium-potential locations that are not near shopping malls, but still require proximity to a hospital, sky train station, or bus station. Ultimately, semi-detached house projects are ideal for medium-potential locations that require proximity to a main road for convenient access to sky train station or public transportation, depending on the specific context.
DOI
10.54028/NJ202423403
First Page
1
Last Page
16
Recommended Citation
Tochaiwat, Kongkoon and Pultawee, Patcharida
(2024)
"House type specification for housing development project using machine learning techniques : a study from Bangkok Metropolitan Region, Thailand,"
NAKHARA (Journal of Environmental Design and Planning): Vol. 23:
No.
1, Article 3.
DOI: 10.54028/NJ202423403
Available at:
https://digital.car.chula.ac.th/nakhara/vol23/iss1/3