This section will explain the logic behind the steps in preparing the feature variables for our model.
- Only the indicators in the dataframe that are truly independent of an indication of fire are selected to contribute to the logistic regression model that will predict the outcome. The reason this is performed is to remove any potential bias in the dataset that may already reveal the outcome of the prediction. This minimizes human interaction with the final outcome. The output of the updated dataframe can be seen in the following screenshot:
Please note that the column Neighborhooods - Analysis of Boundaries is originally ...