Loss Functions and Optimization

Loss Functions

Classification Loss Functions

Binary Cross-Entropy BCE = -[y log(p) + (1-y) log(1-p)]

Categorical Cross-Entropy CCE = -Σᵢ yᵢ log(pᵢ)

Focal Loss - Handles class imbalance

Regression Loss Functions

MSE = (1/n) Σ(yᵢ - ŷᵢ)² - Penalizes large errors MAE = (1/n) Σ|yᵢ - ŷᵢ| - More robust to outliers Huber Loss - Combines MSE and MAE