Data Drift — Google Cloud ML Engineer Practice Questions
Data drift refers to a change in the statistical properties of input features over time, causing a model trained on historical data to make increasingly inaccurate predictions on current data. The Google Cloud ML Engineer exam covers how to detect drift using skew and drift thresholds configured in Vertex AI Model Monitoring, which continuously compares serving traffic distributions against a training baseline. Candidates must understand the difference between feature drift, which affects inputs, and concept drift, which affects the relationship between inputs and outputs, as well as how to respond with retraining or data collection strategies.
Free questions on data drift
Which approach is used to detect when a deployed model's performance degrades due to changes in input data distribution?
Free question · medium · full answer + explanation
More data drift questions in the full bank
- When monitoring production ML models, which metric indicates potential data drift? Unlock answer & explanation →
- Your model serves different geographic regions with varying data. How should you monitor it? Unlock answer & explanation →
- Your continuous ML system detects that model performance has degraded by 5% in production. What is the most likely cause and recommended action? Unlock answer & explanation →