Milot Gashi, researcher of area 3 – Cognitive Decision Support – presented his work at the IEEE Information Visualization (InfoVIS 2020). The IEEE Information Visualization (InfoVis) conference focuses on a diverse set of topics related to information visualization.

He proposes an interactive visual analytics dashboard based on SHAP values that highlights key features in a prediction made by an AI model to predict the fault of complex industrial devices. Furthermore, the dashboard allows domain experts to explore devices with similar behavior in the past.

“Interactive Visual Exploration of Defect Prediction in Industrial Settings through Explainable Models based on SHAP values” was coauthored with colleagues from Pro2Future, Fronius International, AVL, and TU Graz.

#pro2future #InteractiveVisualExploration #DefectPrediction #ExplainableAI #SHAPvalues #DecisionSupportSystem #cognitive

IEEE InfoVIS 2020