Data Science

Evaluating Models: Metrics that Matter

SQL Mastery Team
June 16, 2026
6 min read

It's **Day 146**. You build a model... but is it trash or treasure?

1. Regression Metric: R-Squared (0 to 1)

R-Squared tells you what percentage of the "Variation" in the data your model explains.

  • **0.9**: Legend. Your model is extremely accurate.
  • **0.4**: Okay. You've found a trend, but there's a lot of noise.
  • **0.0**: Fail. Your model is basically a random guess.

    • 2. Classification Metric: Accuracy (%)

    The percentage of correct guesses.

  • "We correctly predicted 95% of the Spam emails."

    • The Trap of Accuracy

    If 99% of your emails are "Not Spam," and your model just guesses "Not Spam" every single time, you have 99% accuracy—but your model is useless! This is why we'll learn about the **Confusion Matrix** tomorrow.

    Your Task for Today

    Calculate the `.score(X_test, y_test)` for your regression and classification models.

    *Day 147: The Confusion Matrix.*

    Ready to put your knowledge into practice?

    Join SQL Mastery and learn through interactive exercises.