PYTHON :

Create functions and loops when needed to avoid manual work

process and analyse one of the datasets available (heart disease, cancer dataset or any dataset available in libraries as sklearn ot tensorflow).

perform data preprocessing (change text to numbers, remove rows with no data, scaling

create at least 3 different feature sets based on correlation analysis or other EDA you have generated

use cross validation for spliting test and training data. generate stats about the samples in both test and training set.

create 4 supervised learning models using KNN, , SVM (there are 2 modesl), NB (there are 3 models) and Random Forest

run the models on each of the three features and save accuracy, precision, f measure and recall of each model in a colum in a dataframe

use the datarafme above to compare the perfromance of models using roc_arc and precions/accuracy

For each classifier compare the result obained by each feature set

For each feature set, compare the result obtained by each classifier

revisit your models, change parameters value and observe the changes in performance

analyse the result and explain why the model performed in a certain way and what are possible option to improve it. (e.g one could have performed better with True Positive and the other with True Negative.

use Ensembling classification algorithm to combine the result of top 5 model.