Using Principle Component Analysis and Clustering Algorithm to Identify Power Factors and Transformation Pattern of Regional Construction

Abstract

The purpose of this research is to analyze the power factors that drive regional development of construction enterprises in order to identify the potential problems within their corresponding transformational pattern. The authors used principal component analysis, and exhaustive clustering experiments and K-means’ ANOVA to verify the feasibility and operability of a regional transformational path. With the data collected from 123 enterprises in the Shaanxi province, China, the results showed that there were 4 power factors: (a) implementation level of transformational development, (b) external information integration, (c) project knowledge management, and (d) corporate brand building support. The overall power factor (Fs) was 6.987, with the σ value of 1.78, min value of 2.895, and max value of 10.843. This research for the first time quantified the influence of the power factors to different clusters of companies and their levels of transformational development. This research discussed their corresponding strengths, weaknesses and provided practical tactics for them to establish suitable development strategies. This research could guide the regional efforts of the industry in reaching higher levels of transformation evolvement, and the emerging framework targets change-based momentum and dynamic clustering elements to the transformational development of regional construction.