🛡️ 04 — Overfitting and Regularization

Neural networks can memorize training data.

---

Signs of Overfitting

• training loss decreases
• validation loss increases
• training accuracy much higher than validation accuracy

---

Regularization Tools

Dropout

layers.Dropout(0.3)

L2 Regularization

layers.Dense(
    32,
    activation="relu",
    kernel_regularizer=keras.regularizers.l2(0.001)
)

Early Stopping

callback = keras.callbacks.EarlyStopping(
    patience=3,
    restore_best_weights=True
)

---

Practical Advice

• start with a small model
• track validation loss
• use early stopping
• scale numeric inputs
• compare against simpler ML models

---

Next

➡️ 05-exercises