A Real Estate Valuation Model Using Boosted Feature Selection

To estimate real estate values, a complex valuation model based on artificial neural network (ANN) has been established as successful means in modern machine learning research, specifically when high-dimensional data are available. Unfortunately, the many locations, such Thailand, quite limited terms of features. Hence, it becomes mandatory to reduce complexity using feature selection techniques. These techniques aim improve performance by identifying significant factors and help decrease computational overload construction. However, due lack explicability interpretability ANNs, analysis input cannot be explained directly composition. In this we apply combination boosting strategy sensitivity an improved Garson's algorithm perform that can adjust its criteria through each iteration ANN model. This proposed technique is then compared with other traditional synthetic real-world house data. The results show our maintain coefficient for every informative feature. study provides set features influences price implies character specific area. It also contributes statistical improvement most cases, lower error than obtained without selected dataset, placed top 24% public leaderboard Zillow Prize competition.