Evaluation of Newton-Raphson Method and Genetic Algorithmfor Estimating Spatial Logistic Model

摘要: This paper evaluates the performance of a mathematical inference method (Newton-Raphson, NR) and an evolutionary computation (Genetic Algorithm, GA) for solving maximum likelihood estimation in logistic analysis. A spatial regression model is formulated to examine relationship between land use various determinants such as population density, distance road, commercial center, etc. Geographic Information System (GIS) used develop data perform The NR GA are utilized estimate coefficients that maximize model. Both methods compared terms computing time. It was found can achieve better also much faster than method. Therefore, recommended although be employed.