The article titled "Adaptive Shapley: Using Explainable AI with Large Datasets to Quantify the Impact of Arbitrary Error Sources" by Birk Magnussen, Maik Jessulat, Claudius Stern and Bernhard Sick presents a method that uses continuous feature networks to calculate Shapley values for individual sensor channels and reveal their contribution to production-related measurement errors. The approach is demonstrated on optical sensors in spatially resolved reflectance spectroscopy and provides insights for production and quality control by correlating sensor quality metrics with measurement accuracy.

The article titled "Adaptive Shapley: Using Explainable AI with Large Datasets to Quantify the Impact of Arbitrary Error Sources" by Birk Magnussen, Maik Jessulat, Claudius Stern and Bernhard Sick presents a method that uses continuous feature networks to calculate Shapley values for individual sensor channels and reveal their contribution to production-related measurement errors. The approach is demonstrated on optical sensors in spatially resolved reflectance spectroscopy and provides insights for production and quality control by correlating sensor quality metrics with measurement accuracy.