Document Type : Research Paper
Authors
1 Professor of the Faculty of Geography, University of Tehran
2 Ph.D. Student of Urban Planning
Abstract
Until now, a major application of scientific methods has been to estimate and predict the values of a variable by referring to the values of one or more other variables associated with it. In this paper, we examine the most common type of "correlation", or estimation based on structural relationships between variables, which leads to the creation of a simple linear model.
The use of linear models in a variety of urban analyses is quite commonly accepted due to the relative simplicity and accuracy of these models. Linear models usually show tendencies over time, which may not be taken into account through abstract or chart-based analyses. Linear models are used to express events when non-scientific methods are incomplete and unsatisfactory (and predict the appearance and time of the events).
Keywords