saberrg / credit-risk-classification

Machine Learning Model that Evaluate the risk of given loans based on historical lending activity

Credit Risk Classifications

Machine Learning Model that evaluates the risk of loans based on historical lending activity. Determines whether loans are healthy (0) or high risk (1).

Overview

The purpose of this analysis is to use the data in "lending_data.csv" to train a logistic regression classifier then having it predict make predictions. An example of a supervised machine learning. Then the predication were examined to see how accuracte the model was making predictions on healthy and high risk loans.

The Accuracy, Precision and Recall Scores

• For class 0, both precision and recall are 1.00, indicating that it performed exceptionally well in identifying class 0.
• For class 1, while the precision is slightly lower (0.87), the recall is still high (1.00), suggesting that the model correctly identified most instances of class 1 while being somewhat conservative in making positive predictions.

Summary Conclusion

According to the accuracy score the model averaged a 94%, which gives a high level of prediction accuracy. Since this was done twice and both times it received high scores in accuracy. I would highly recommend this model to the company for predicting which loans are safe and which aren't so.