Metrics to Evaluate your Machine Learning Algorithm: Accuracy, Precision, Recall, Specificity, and F1.

This article will discuss the most common ML model evaluation metrics such as Accuracy, Precision, Recall, Specificity, and F1 Score for a classification problem in the fintech space.

This article is part of a series where we walk step by step through solving fintech problems with different Machine Learning techniques using the “All lending club loan” dataset. Here you can find the complete end-to-end data science project for beginners to learn data science.

Introduction to Data Science with python: A complete guide to learn data science and machine…

We have walked through the confusion matrix in the previous article, and I suggest you start from there.

Drawing by author

If you are preparing for an interview, this article would help you to answer the following questions:

1. What are precision and recall?
2. What error metric would you use to evaluate how good a binary classifier is?
3. How do you find the accuracy of a confusion matrix?
4. What is F1 score in confusion matrix?
5. What does 1% accuracy mean?
6. What is the difference between precision and accuracy?
7. In what cases you should not use accuracy as the main metric?
8. What is Specificity?
9. When should you use precision, and recall?
10. What is a Negative predictive value?

The confusion matrix helps us visualize whether the model is “mistaken” in distinguishing between two classes. As you can see in the below picture, it is a 2x2 matrix. The row names are the actuals from the test set, and the column names are the ones predicted by the model.