Use different deep-learning method to improve the performance of the coronavirus test kit in terms of recall and precision.
*The test kit can automatically take 25 different measurements. The reading of the first measurement x1 is categorical. The rest of the readings x2 to x25 are numerical.
Use different classifiers included ExtraTrees, MLP, SVC, kNeighbors, RandomForest, soft-voting to get best hyperparameters that training more than 4000 samples that achieve 100% recall and more than 70% of precision.