##Response
The scientists use a Random Forest estimation technique, which involves setting up covariate boundaries and then taking and supplementing data. After that, the data is rasterized and projected so that it can be revised and so covariates can be created. Finally, the algorithm then estimates the random forest model and predicts the per-pixel population density and eventually makes it more specific.
A machine learning algorithm is an algorithm that uses patterns in data and inference to do something or make predictions. It is different from normal statiscal approaches since it is capable of coming to a conclusion about data itself while also still obtaining data. It stil self-sufficient, and only needs a few things(in the random forest estimation technique’s case) to work.
Some of the covariates used were related to different kinds of land cover, a census of the population, lights at night, distance of roads, size of bodies of water, and different towns, cities, and villages.The boundaries were set to the country borders, and then buffered by 10 kilometers. Because the data in big data is so large, machine learning methods are much more effective when combined with big data because computers can already process information faster than humans can. By setting up a algorithm that lets computers automatically infer and sense patterns in data to make conclusion, less human work has to be done and more time can be saved.
There are multiple benefits to knowing the locations of people, especially those in LMICs. It is possible to see patterns in many different aspects of life, such as seeing migration routes or travel routes to see where people may congregate to or see a location where people believe they can get better jobs or pay. In natural disaster times, it is also better to know the location of people as it makes providing aid and saving lives much easier. In terms of disease vectors, knowing common travel paths and knowing the populations of certain cities makes it so that preventing infection is much more effective.
Migration in China is on a massive scale and the numbers of people traveling, especially around certain holidays like Chinese New Year. Knowing locations allow patterns such as socioeconomic differences and where poor and rich people migrate to. It also allows us to see the travel patterns of people not only on holidays, but also just on a normal day. We can also tell if more people are moving from rural to urban areas, and how many of those people stay, leading to possible patterns of overcrowding and pollution in cities. Knowing all this info allows governments to better prepare urban planning and work more efficiently.