Loss Functions Checkpoint

1. Mean Squared Error is best for:

Regression tasks Classification tasks Clustering Dimensionality reduction

2. Cross-entropy loss is commonly used for:

Classification Regression Clustering Feature selection

3. A lower loss value typically means:

Better model predictions Worse model predictions More overfitting Less training data

4. What function is often applied before cross-entropy in classification?

5. Binary cross-entropy loss is:

-[y*log(p) + (1-y)*log(1-p)] (y - p)^2 |y - p| y * p