StackedML
Practice
Labs
Questions
Models
Pricing
Sign in
Questions
/
Model Evaluation & Experimentation
/
Validation
/
Time-series CV
← Previous
Next →
812.
Walk-Forward Validation Definition
easy
What is walk-forward validation in the context of time series cross-validation?
A
Splitting the time series into equal-length blocks and rotating which block serves as the validation set
B
Training on the full historical dataset and validating on multiple non-overlapping future windows simultaneously
C
Training on a growing or rolling window of past data and validating on the immediately following period
D
Alternating between training and validation folds to ensure each time period appears in both roles
Sign in to verify your answer
← Back to Questions