General Diagram of the Project
Factors relevancy

Clinical-microbiological characterisation of SARS-CoV-2 infection in the paediatric age

Group: Pana Baya

Members:

Marc Garcia
Jofre Poch
Pau Tarragó
Pau Matas
Tomás Gadea

General description

Our project predicts the diagnosis of children using both probability and machine learning. Firstly, we select the impactful variables we want to consider in our predictions using a covariance test between each variable and the final diagnosis.

Afterwards our machine learning model is trained based on the important features. The model classifies each input as either COVID-19 positive or negative. The chosen model is a random forest classifier, which fits the most our data. After this process we are ready to predict with relatively high accuracy the diagnosis of a child.

The program takes an input file and through our predictor returns a file with the diagnosis of each case.