The technique described by the authors (Stevens, et.al), is called the “Random Forest” estimation technique. It is a way to estimate population density within countries by taking in census data and ancillary data. Random forest allows for multiple learning methods to be brought together by creating “trees”, and then combining them to create accurate predictions. The technique is used to create more accurate and flexible gridden predictions as compared to other methods previously used.
A machine learning algorithm is a way to get computers/technology to automatically complete a task by interpreting patterns without having to provide step by step instructions. This method is different from others as it allows for lots of data to be gathered and interpreted with less effort, and allows for more accurate predictions to be made.
The covariates mentioned are vector and raster class data such as land coverage, bodies of water, and the absence or presence of protected areas. They represent a big portion of the data that was studied, and all of the big data plays an extremely important role in the machine learning methods as it offers large amounts of information to study and learn from. This studied data in turn will allow for more accurate predictions of the human population distributions.
By having access to accurate data on people’s locations, many opportunities arise to help the planet. We will have more knowledge on under-developed countries and how to support them, locations of diseases and how they spread, how to accommodate more densely populated areas, post-crisis assistance, and much more.
In accordance to my research topic, having accessible population data is quite important. I am researching the effects of the DTP vaccination and it is extremely important to know who has received the vaccination versus those who have not. If accurate data came out to show the exact locations of where people are more densely populated, we could then determine which areas would need more vaccination assistance versus others. Stopping the harmful effects of these diseases is extremely important, and this data could be used to eradicate these diseases from planet Earth.