An ensemble learning based model for real estate project classification

摘要

The demand of accommodation has been increasing especially in the metro area. The process of credit assessment for loan is increasingly high. One of the criteria for banks to determine the amount of loan is a grade of the real estate project. This paper addresses the problem of determining the grade of a real estate project and proposes a real estate classification model. The model helps loaners to make a decision for the further stage of the loan. For the model construction, the data were gathered from 407 real estate projects in Thailand which are launched in 2014 and forthcoming projects that will be launched in 2015. The variables of our model involve the project infrastructure and characteristics such as the facilities, number of units, location, parking space, and size of a developer. For all 407 projects, the grades were provided by a bank. We use the supervised learning with ensemble technique and vote algorithm for training and testing against the dataset where the dataset is separated into the training set and the testing set of 307 records and 100 records, respectively. The efficiency of our proposed model is measured by classification accuracy and user satisfaction with the model.