Prediction of Diabetes Mortality in Mexico City Applying Data Science

Around the world, diabetes is a disease that in recent years has had a significant increase in mortality rates. Currently, several countries consider diabetes an important public health problem, particularly Mexico. In this research work, the problem of mortality forecasting for the next 5 years in Mexico City was addressed by applying Data Science. For developing the application, an extension of the IBM methodology called Batch MFCD was used, which is oriented to the epidemiology domain. All the mortality and other data used in the research are from public and official databases and belong to the 1990-2019 interval. For forecasting we selected the Support Vector Regression (SVR) model, which allowed forecasting the mortality rates for Mexico City in the 2020-2024 interval. A decrease of the mortality rate for the 2017-2019 interval was observed for the actual data, and for the 2020-2024 interval it is forecasted that mortality will continue to decrease at a similar rate. It is worth mentioning that 2020 was an atypical year, because of the COVID-19 pandemic; therefore, it is foreseeable that its effect may affect the actual mortality rates in subsequent years.