From the course: Data-Centric AI: Best Practices, Responsible AI, and More
Join today to access over 24,700 courses taught by industry experts.
Understanding data drift and model drift
From the course: Data-Centric AI: Best Practices, Responsible AI, and More
Understanding data drift and model drift
- [Instructor] Now, let's dig deeper into each of the causes of data drift and how we can ensure that we are using the right means to measure and correct them. Real world data is not static. Rather, it evolves gradually over time in response to changing human behaviors, new technologies, economic trends, regulations, and other external factors. Population demographics and characteristics tend to change slowly as the cultural shift occurs. New categories and types of data emerge as behaviors and technology progresses. The processes generating the data can also drift, causing distribution changes. For example, how customers interact with a company's website may change as a new feature is added or removed. Major external events like recessions, innovation, or new laws can significantly disrupt existing data patterns. We all remember what happened during the COVID-19 pandemic. With changing economic climate and day-to-day events,…
Download courses and learn on the go
Watch courses on your mobile device without an internet connection. Download courses using your iOS or Android LinkedIn Learning app.