Application of Improved Logarithm Logistic Models in Population Prediction

Abstract

In this paper, improved Logistic models are given, which are called Logarithm Logistic Models. Based on U.S. Census data, the parameters of the models were estimated by applying the Least Squares Method. The experiments show that the prediction value of the new models is much closer to the actual value than the classical Logistic model. Finally, through analyzing the rationality of the maximum population capacity, the trend of Logistic curves and the rationality of prediction value, the most appropriate model is recommended.