IGWO-SVM: An Enhanced Cost Prediction Model for Construction Projects

Abstract

Accurate cost prediction is crucial for successful construction project management. This study proposes an improved Grey Wolf Optimizer-Support Vector Machine (IGWO-SVM) model for construction project cost prediction. The model incorporates Tent mapping and quantum well techniques to enhance the global search capability and prediction accuracy. Experimental results show that the IGWO-SVM model achieves a prediction error rate of 0.01%, significantly outperforming traditional methods. The model reduced total construction days by 1.72%, total cost by 1.89%, and improved quality levels by 15.31% on average. The IGWO-SVM model demonstrates high stability and accuracy, providing a reliable tool for construction project cost prediction and optimization.