Credit_Risk_Analysis

Overview of the Loan Prediction Risk Analysis

Application of machine learning to evaluate credit card risk by employing different techniques to train and evaluate models.

Results of the Six Machine Learning Models

Random Over Sampler

Balanced Accuracy Score: 65% Precision Score: 1% Recall score: 62%

Smote

Balanced Accuracy Score: 64% Precision Score: 1% Recall score: 63%

Cluster Centroids

Balanced Accuracy Score: 52% Precision Score:- 1% Recall score:- 61%

Smoteenn

Balanced Accuracy Score: 63% Precision Score: 1% Recall score: 71%

Balanced Random Forest Classifier

Balanced Accuracy Score: 79% Precision Score: 3% Recall score: 70%

Easy Ensemble Classifier

Balanced Accuracy Score: 93% Precision Score: 9% Recall score: 92%

Summary of the Results

Based on the abaove analysis I would recommend using the Easy Ensemble Classifier Model to determine the credit risks as it shows the highest accuracy score of 93%.