The authors use a multi-stage technique called “Random Forest estimation technique”. This technique first compiled the data from important covariates and log population density into a TuneRF function, which helps focus in the “Random Forest estimation technique”. The dependent variable, population density, was influenced by predictors. Essentially, it is easy to think about it as a tree, and each piece of data is sent through the variables to further classify it into more specific data. A machine learning algorithm is something that can help predict reasonable and acceptable range. So these algorithms are good at accurately predicting, or predicting as accurately as possible. Within this study, the data was able to be better “predicted” by using this new technique, rather than the old way of using purely statistics. By using the data science methods recently created, that include using geospatial covariates, we can begin to explore covariates while processing data, making it more accurate in the long run. We are able to define variables and see how they react with other variables. In this case, they used a number of land and topographical related variables in order to specify their data. They used a number of variables as their covariates, in order to improve these predictions. This, and more, are considered predictors. They used shrubs, bare areas, water bodies, and other land related variables to help see if they could pinpoint exactly where each person and family resides. This makes it much more specific. Big data gives even more variables to send data through, and adds even more “layers” to the geospatial framework, so that data is more accurate.
In my research, these sorts of geospatial covariates could be used to figure out a better estimation of where the more women reside regarding child-bearing age. If this result was more accurate, then we could better provide and allocate resources to those who need them most. For instance, right now, Ethiopia has several healthcare buildings, but women would have to walk for hours, or even days to get to where they need to go.