Autoencoder-Enabled Potential Buyer Identification and Purchase Intention Model of Vacation Homes

A trend of purchasing a lakeside, seaside, or forest vacation home has been raised in China. However, such purchase behavior has received limited attention from the research community in emerging markets. This study aims at investigating the factors behind vacation home purchase behavior and helping identify potential buyers. Specifically, factors, such as air quality, enduring involvement, place attachment, and destination familiarity, are examined via a proposed integrative model, which links these factors to purchase intention. The total number of potential buyers of vacation homes is increasing but remains small, compared to the whole consumers' population, resulting in imbalanced purchase behavior data when validating a model. To address this problem, this study proposes an autoencoder-enabled and k-means clustering-based (AKMC) method to identify potential buyers. The proposed methods tested on a dataset of 309 samples, collected through a questionnaire-based survey, and achieves a model accuracy of 82% in identifying potential buyers, outperforming other traditional machine learning methods, such as decision trees and support vector machines. This study also provides explainable results for the vacation home purchase behavior and a decision-making tool to identify potential buyers.