Logistic Regression In Modelling Some Sustainable Development Phenomena

Technological innovations of the last decade have led to a real explosion of data and a practically unlimited capacity to create and to store them, remodelling day to day life. This paper analyses theoretical models for qualitative variables used in sustainable development. More exact and detailed information on natural resources are vital to the state, as well as to environmental agencies and to the private sector. The type of forest vegetation is one of the basic characteristics that are recorded and analysed in order to maintain the ecological balance. Generally, the type of forest vegetation is either recorded directly by the agents, or by tele-detection. Both techniques are costly both in financial and time terms or even impossible to do. Predictive models offer an alternative to obtain this data. Although linear regression models are used on a wide scale by biologists and ecologists, these models are inadequate when the dependent variable is qualitative.. Logit models are a natural complement to regression models, where the endogenous variable is a qualitative variable, a situation that may be obtained or not, or a category of a classification. The popularity of logit models is explained by the multivariate nature of the models and the easiness with which they can be interpreted.