Evolving Fuzzy Systems Based on the eTS Learning Algorithm for the Valuation of Residential Premises

摘要

An attempt has been made to employ evolving Takagi-Sugeno algorithm (eTS) to built models assisting property valuation on the basis of actual data drawn from cadastral system, registry of sales transactions, and a cadastral map. Seven methods of feature selection were applied an evaluated. The eTS performance was compared to three algorithms implemented in KEEL, including decision trees for regression, neural network, and support vector machine. The results confirmed the advantages of the eTS algorithm.