Computing green remodeling construction cost for public buildings based on genetic algorithm and case-based reasoning

As a first step in considering the budget limitations caused by public funds, a systematic method for estimating the green remodeling construction costs during the planning stage must be established first. In this research, we developed models to predict the construction costs of green remodeling by applying case-based reasoning and genetic algorithm. The applicability of the models was experimentally evaluated, wherein 15 datasets were randomly selected to validate the estimated preliminary construction costs. The average similarities of the retrieved cases were higher than 97 %. The overall performances of the models were evaluated, and the mean absolute error rates for 1-, 3-, 5-, and 10-nearest neighbors were 10.38 %, 7.15 %, 6.91 %, and 9.45 %, respectively. This research contributes to literature of building retrofit cost management in construction by not only offering the systematic approach for estimating the preliminary construction cost, but also providing as foundational data when deriving budgeting strategies.