A model validation and predicting the rental values of residential properties using logistic regression model

The property market is a key contributor to the economic growth of many countries. This makes information from property valuation reports vital for decisions on real estate investments and property tax. Unfortunately, literature reveals that inaccurate property valuation arising from a reliance on traditional methods of valuation remains a major problem facing real estate practice. To improve the prediction accuracy of property valuation estimates, modelling techniques such as neural networks have previously bejgn applied to this problem. This present study uses a logistic regression model to predict the rental values of residential properties in Cape Town, South Africa. Field survey data was divided into two groups: training and test sets. The training set was used for model development while the test set was used for model validation. The results of the study revealed that parking, garden, number of bedrooms and floor area have the most significant impact on the rental values of residential properties. Surprisingly, proximity to a police station has one of the least effects on the rental values of residential properties. With a prediction performance of over 70% accuracy, findings indicate that the logistic regression model is suitable for predicting the rental values of residential properties. This study evaluates the factors that influence the rental values of residential properties located within the study area. The developed model can serve as a decision support tool for estimating the tax payable by property owners.