nerdexam
Google

PROFESSIONAL-MACHINE-LEARNING-ENGINEER · Question #84

You work for a global footwear retailer and need to predict when an item will be out of stock based on historical inventory data Customer behavior is highly dynamic since footwear demand is influenced

The correct answer is C. Use the last relevant week of data as a validation set to ensure that your model is performing. For time-series inventory prediction with dynamic behavior, using the most recent relevant data as a validation set is the most streamlined and reliable way to ensure a model's current performance before deployment.

Submitted by asante_acc· Apr 18, 2026ML model development

Question

You work for a global footwear retailer and need to predict when an item will be out of stock based on historical inventory data Customer behavior is highly dynamic since footwear demand is influenced by many different factors. You want to serve models that are trained on all available data, but track your performance on specific subsets of data before pushing to production. What is the most streamlined and reliable way to perform this validation?

Options

  • AUse then TFX ModelValidator tools to specify performance metrics for production readiness.
  • BUse k-fold cross-validation as a validation strategy to ensure that your model is ready for
  • CUse the last relevant week of data as a validation set to ensure that your model is performing
  • DUse the entire dataset and treat the area under the receiver operating characteristics curve (AUC

How the community answered

(35 responses)
  • A
    23% (8)
  • B
    6% (2)
  • C
    60% (21)
  • D
    11% (4)

Why each option

For time-series inventory prediction with dynamic behavior, using the most recent relevant data as a validation set is the most streamlined and reliable way to ensure a model's current performance before deployment.

AUse then TFX ModelValidator tools to specify performance metrics for production readiness.

TFX ModelValidator tools are excellent for defining and enforcing performance metrics, but the choice of *which* validation data to use (the strategy) is a more fundamental consideration for dynamic time-series data.

BUse k-fold cross-validation as a validation strategy to ensure that your model is ready for

K-fold cross-validation is inappropriate for time-series data as it breaks temporal dependencies and can lead to data leakage by training on future data to predict past data.

CUse the last relevant week of data as a validation set to ensure that your model is performingCorrect

For time-series data with dynamic customer behavior, using the most recent relevant period (e.g., the last week) as a validation set is crucial. This approach simulates how the model will perform on new, unseen data that reflects current trends, ensuring its relevance for immediate production use.

DUse the entire dataset and treat the area under the receiver operating characteristics curve (AUC

Using the entire dataset as a validation set would lead to data leakage and an overoptimistic evaluation, as the model would be tested on data it was trained on.

Concept tested: Time-series model validation strategies

Source: https://cloud.google.com/blog/products/ai-machine-learning/building-a-machine-learning-model-for-time-series-forecasting

Topics

#Model Validation Strategy#Time-series Data#Production Readiness#Data Splitting

Community Discussion

No community discussion yet for this question.

Sign up to join the discussion
Full PROFESSIONAL-MACHINE-LEARNING-ENGINEER Practice