Construction Firms' Operation Management Performance Measurement Model Analysis Using Artificial Neural Network

In an age valuing competitiveness, artificial neural network is widely used to construct measurement model of organizational efficiency in every industry or trade. The researcher uses artificial neural network as measurement model of construction companies' organizational efficiency. After summarizing references and interviews with construction executives of five listed or OTC-listed companies of more than ten years old, the researcher concludes that organizational efficiency of construction companies is classified into five categories and 17 variables and develops 17 measurement items. For this research, 200 questionnaires have been handed out and 195 questionnaires have been collected. The 135 effective samples include 105 learning samples and 30 prediction samples. Through traditional back-propagation artificial neural network simulation and that of variable learning efficiency, it is found that organizational efficiency of construction companies should be classified into four categories, namely, background of organizational structure; flexibility, rules and regulations of organizations; adaptation process of personnel; and methods and objectives of organizational strategy. It is also found that back-propagation artificial neural network of variable learning speed is an excellent prediction tool for organizational efficiency construction applied in this research if over fitness can be avoided.