🟢 Repeat: Bias-Variance (Can You Still Explain It?)

Without looking at Module 05, answer:

1. What are the two components of reducible error?
2. Which one means "too simple"?
3. Which one means "memorized noise"?
4. Your model: 98% train accuracy, 65% test accuracy. What's wrong?

Answers: 1. Bias and variance. 2. High bias = too simple (underfitting). 3. High variance = memorized noise (overfitting). 4. Overfitting — massive gap between train and test. Regularize, get more data, or simplify model.