Predicting Threatened Habitats Using Logistic Regression Based on Environmental Factors

Habitat destruction is one of the most acute problems in the world, which is governed by several environmental features: deforestation, pollution, and climate change. Classic conservation approaches predict the degradation of habitats that result from the complex relationships among these factors often less than perfectly. Logistic regression is used in this research to model the probability of threatened habitats, considering major environmental factors pertinent to the rates of deforestation, CO2 emissions, air and water quality, urban population growth, and temperature increase. In this respect, logistic regression is employed here because it’s simple yet gives highly interpretable output, thus sitting among the most popular methods which are applied for binary classification problems on ecological systems. This study models the impact of those environmental variables in pursuit of better decision-making processes for conservation planning, providing active measures that could be undertaken to mitigate habitat threats. Further, this study shows the importance of understanding the relationship of such predictors with habitat health, as it contributes to the development of more sophisticated conservation strategies.