StackedML
Practice
Labs
Questions
Models
Pricing
Sign in
Questions
/
ML Fundamentals
/
Core Concepts
/
Train / validation / test separation
← Previous
Next →
636.
Purpose of Validation Set
easy
What is the purpose of the validation set in a machine learning pipeline?
A
To measure final unbiased model performance after all training and tuning decisions are complete on unseen evaluation data
B
To detect data leakage by comparing the distribution of training and test set features
C
To augment the training set when labeled data is scarce by providing additional supervised examples
D
To tune hyperparameters and make model selection decisions during development without touching the test set
Sign in to verify your answer
← Back to Questions