The name of the technique is called Random Forest. Random Forest is essentially an estimation technique used to generate a gridded prediction of population density from the census and geospatial data. It works through “bagging” - a term that references combining different learning models to increase classification accuracy. A machine learning alogrithm is different from a regular algorithm because it adjusts itself to become more accurate and precise after being exposed to more data. A random forest model adapts to different input. It is different from classical satistical approaches like describing and analzying phenomenon and events because the machine learning algorithm can take into account any potential changed in covariates or facotrs that may influence the product produced. It is important to have an accurate distribution of population distirbution for various reasons in LMICs. One of these could be for NGOs to deliver aid to people - they must know exactly where people are. Furthermore, in order to fully accomplish the sustainable development goals with the idea that no one gets left behind, it is important to know exactly where everyone is. Thus, one may measure if the success of a program that attempts to fulfill a sustainable development goal. Within the context of my own research, this algorithm would be very helpful because information population distirbution is crucial to epidemiology. If I know where people are, I could implement programs or policies to limit the spread of a certain pathogen. Additionally, it would be beneficial information for delivering medicines or setting up health clinics, so that people get more access to health care facilities.