massOfai

Validation

AI/ML Fundamentals

Evaluating model performance during training

What is Validation?

Using a separate dataset to tune model parameters and prevent overfitting. Like practice tests before the final exam.

Real-World Examples

  • •Splitting data 60% training, 20% validation, 20% test
  • •Cross-validation for small datasets

When to Use This

To tune hyperparameters and select best model

Related Terms

Learn more about concepts related to Validation

Training

Process of teaching an ML model using data

Testing

Final evaluation on completely unseen data

Cross-Validation

Validating on multiple data splits

Overfitting

Model learns training data too well, performs poorly on new data