Loss Functions

Back to Neural Network Fundamentals

Functions that measure how far the model’s predictions are from the true values. The loss is what backpropagation minimizes during training.

Key Types

  • Cross-Entropy — standard for classification (binary and multi-class)
  • MSE (Mean Squared Error) — standard for regression
  • Hinge Loss — used in SVMs, margin-based
  • Contrastive Loss — for learning embeddings and similarity

deep-learning loss-functions training