New Efficient Regression Method for Local AADT Estimation via SCAD Variable Selection

This paper focuses on the estimation and variable selection for the local annual average daily traffic (AADT). The variable selection procedure by smoothly clipped absolute deviation penalty is proposed. It can simultaneously select significant variables and estimate unknown regression coefficients in one step. The estimation algorithm and the tuning parameters selection are presented. The data from Mecklenburg County, North Carolina, USA, in 2007 are used for demonstration with our proposed variable selection procedures. The results show that this penalized regression technology improves the local AADT estimation along with satellite information, and it outperforms some other benchmark models.