Prediction of the performance related to financial capabilities using multilayer perceptron

Abstract

In construction projects, there are circumstances when contractors meet financial prequalification requirements but show low financial performance in practice. These cases bring about the complexity in contractor selection. Hence, the aim of this research is to build a prediction model that finds contractors’ financial performance to support decision makers assess contractors more efficiently in prequalification phase. Thus, this study takes recent roadwork Term Contracts Projects with each with the corresponding contractor's records to train the model to predict Performance related to Financial Capabilities PFC. The Multilayer Perceptron MLP is utilized to find the nonlinear correlation between the PFC and contractors' characteristics. The research finds that more financialcompetitive contractors show less financial performance than less competitive ones. The findings of the research help the client improve the current contractors’ evaluation system to exhaust the possibilities of financial performance.