Question: What does \'model drift\' refer to in the context of deployed models?

Options:

1. The process of updating the model with new data
2. The degradation of model performance over time due to changes in data distribution
3. The initial training phase of the model
4. The difference between training and testing datasets

Correct Answer: The degradation of model performance over time due to changes in data distribution

Solution:

Model drift occurs when the statistical properties of the target variable change over time, leading to decreased model performance.