Leakage Detection Challenges
Sarah discusses the complexities of leakage detection in predictive modeling, particularly in churn prediction. She highlights the importance of identifying potential issues, such as using the reason for churn as a feature, which can lead to performance degradation over time. The conversation emphasizes the need for continuous monitoring and innovation in feature engineering to maintain model accuracy.In this clip
From this podcast
The TWIML AI Podcast (formerly This Week in Machine Learning & Artificial Intelligence)
Agile Data Science with Sarah Aerni - #143
Related Questions