Open CynthiaLiu0805 opened 1 month ago
This model demonstrated a correlation coefficient of 0.76. Considering the probable errors in the measurement of all three factors, the determination factor is considered to be remarkably good. It is likely that the regression could be improved substantially if the composite chloride profiles were to be based on a much larger sampling of each deck, and if snowfall and traffic data for the actual years of service could be used. (Weyers, R. E., M. G. Fitch, E. P. Larsen, I. L. Al-Qadi, W. P. Chamberlin, and P. C. Hoffman. 1994. Concrete bridge protection and rehabilitation: Chemical and physical techniques. Service life estimates. Rep. No. SHRP-S-668. Washington, DC: National Research Council.)
Please review the IM2 of 4.4.5 Instance Models section of the SRS, how is this linear regression reliable and how possible would it change in the future?