How can water systems effectively classify unknown service lines using predictive modeling? In this video from the series, “Expert insights: New standards for water infrastructure,” BlueConduit’s VP of Data Science, Alice Berners-Lee, joins VP of Sales, Kristy McGrath, to explore how water systems and their engineering partners can leverage predictive modeling to classify unknown service lines as lead or non-lead.
🔍 Key topics covered:
✔️ Understanding the concept of unknown management in lead service line inventories
✔️ How predictive modeling assigns lead likelihood probabilities to unknown service lines
✔️ The role of thresholds in classifying service lines and how they vary by state
✔️ The importance of calibration, metrics like precision and recall, and confidence intervals in decision-making
✔️ How water systems can use data-driven insights to guide replacement planning and compliance
BlueConduit has spent nearly a decade developing best-in-class predictive modeling tools to help water systems make informed, high-quality decisions. If you’d like to learn more about reducing unknowns in your inventory, reach out to our team.