DataCamp

Part of

← Data Science

Datacamp

• ((f3966dc1-0f09-4704-9f13-0a574792e917))

  • Supervised Learning with Scikit-Learn

    • Regularization

      • Lasso is great for feature selection

        • Lasso performs regularization by adding to the loss function a penalty of the aboslute value of each coefficient multiplied by some alpha.

        • Also known as L1 regularization because the regularization term is the L1 norm of the coefficients.

      • When building regression models, ridge regression should be your first choice.

        • taking the sum of the squared values of the coefficients multiplied by some alpha

        • L2 norm

    • Diagnosing classification predictions

    • Confusion Matrix

      • • Predicted: Spam Email

          • Predicted Real Email

        • True Positive

          • False Negative

        • False Positive

          • True Negative

      • Accuracy

        • $\frac{tp+tn}{tp+tn+fp+fn}$

      • Precision

        • $\frac{tp}{tp+fp}$

      • F1 score: $$$$