Done by Bibarys Mussagaliyev, Department of Electrical and Computer Engineering, Nazarbayev University
Project is about predicting the bank-loan approval based on the different features of individuals. Project's features: I have used four classifiers: Logistic regression, Random forest, XGBoost, kNN, in addition: feature selection based on Sequential Forward Search was implemented. Models were evaluated using confusion matrices, AUC ROC, CV-accuracies. Project was done on Python3, and I used Jupyter Notebook as development environment. Run the notebook by Anaconda. Project contains two .csv files (test and train data sets). Succesful run of project will output the .csv file (predicted values).
Regards, Bibarys!