What is model drift in predictive analytics and why does it occur over time? How does model drift affect the accuracy and reliability of machine learning models? What are the different types of model drift, such as data drift and concept drift? How can organizations detect and monitor model drift in production systems? What strategies can be used to handle and reduce model drift in predictive analytics models?