Data drift is best described as a change in data distribution over time affecting model performance.

Data drift is best described as a change in data distribution over time affecting model performance.

• A. False
• B. True
• C. It always improves accuracy
• D. It is identical to data leakage

Answer: (B)

Data drift happens when the input data’s statistical properties change over time compared with what the model was trained on. That shift can make predictions less accurate because the model is operating on data that no longer matches its learned patterns. So describing data drift as a change in data distribution over time that affects model performance is accurate. It doesn’t guarantee improved accuracy—drift often leads to worse performance. It’s also not the same as data leakage, which involves exposing the model to information during training or evaluation that shouldn’t be available, causing unrealistic performance estimates.

Data drift happens when the input data’s statistical properties change over time compared with what the model was trained on. That shift can make predictions less accurate because the model is operating on data that no longer matches its learned patterns. So describing data drift as a change in data distribution over time that affects model performance is accurate. It doesn’t guarantee improved accuracy—drift often leads to worse performance. It’s also not the same as data leakage, which involves exposing the model to information during training or evaluation that shouldn’t be available, causing unrealistic performance estimates.