Cold season has become a major public health problem in Mexico. Only in 2018, there were more than 23.8 million cases of acute respiratory infections and 500 deaths due to influenza.
How could we make Mexicans understand the risk of the cold season?
We created the most accurate predictive model, that estimated the risk of flu all over Mexico in real time and was displayed on web banners. To make this possible, we leveraged data from the National Institute of Epidemiology, analyzing more than 8 years of infection cases. Then we layered in weather data from more than 80 cities considering over 50 variables, such as temperature, weather, humidity, etc. Finally we included the analysis of keywords trends in Google searches, and geolocated social media activity related to flu symptoms.