A Method for Improvement the Parameter Estimation of Non-linear Regression in Growth Model to Predict Project Cost at Completion

摘要

In this paper, we propose a comparison between existing parameter estimation methods of a nonlinear regression-based growth model to forecast project duration as well as its cost at completion. To analyze and perform some experiments, we used the Gompertz growth model and the dataset comes from a number of similar previous studies. Two other nonlinear models were also applied to compare the results with the Gompertz model in terms of fitness-function score. The performance of the proposed methods is also the key for further studies in fitting the S- curve and predicting the Estimate-to-Complete and the Estimate-at-Completion of certain projects.