Bidding Price Forecast of Construction Projects Based on Machine Learning Algorithm
Haixia Li · 2022
In this study, we focus on the need for the parameters of the Support Vector Machine (SVM) which is optimized by the Particle Swarm Optimization (PSO) algorithm. SVM has higher recognition advantages in solving nonlinear problems for small samples and high dimension pattern problems. By choosing the Gauss longitudinal kernel function as the kernel function of the nonlinear decision function, the nonlinear problem is converted to a linear problem by using SVM. A detailed presentation of the required kernel function is defined. Under the premise of guaranteeing the parameter optimization of SVM, PSO algorithm for the optimal speed improves the computing speed of SVM. Through Matlab2014b, the results show that the bidding price prediction strategies based on the PSO-SVM algorithm have significant feasibility.