Evolutionary fuzzy decision model for construction management using support vector machine

Construction projects are, by their very nature, challenging; and project decision makers must work successfully within an environment that is frequently complex and fraught with uncertainty. As many decisions must be made intuitively based on limited information, successful decision making depends heavily on two factors, including the experience of the expert(s) involved and the quality of knowledge accumulated from previous experience. Knowledge, however, is subject to various factors that cause its value and accuracy to deteriorate. Research has demonstrated that artificial intelligence has the potential to overcome these factors. The Evolutionary Fuzzy Support Vector Machine Inference Model (EFSIM), an artificial intelligence hybrid system that fuses together fuzzy logic (FL), a support vector machine (SVM) and fast messy genetic algorithm (fmGA), represents an alternative approach to retaining and utilizing experiential knowledge. A fmGA is used as an optimization tool to search simultaneously for fittest membership functions, defuzzification parameter (dfp) and SVM hyperparameter (herein C and gamma, γ). Two simulations on actual construction management problems demonstrated the EFSIM to be an effective tool for solving various problems in the construction industry.