Classifier + bayesian optimization for COVID-19 prediction

Albert Einsten Hospital made COVID-19 data available on Kaggle, the tasks are to predict whether the patient is infected and what type of hospital bed it should be referred. The predictive variables are tests such as blood count, urine test, among others.

I submitted a solution to task 1, which is to predict whether the patient is infected or not, I tried three classification models, Random Forest, XGBoost and Multi-layer Perceptron with Bayesian optimization of hyperparameters.

link: https://www.kaggle.com/plbalmeida/task-1-classifier-bayesian-optimization