Qlik Predict - Understanding Coordinate SHAP Analytics

In this installment of the Do More with Qlik Tips and Tricks series, Mike Tarallo guides you through creating visual analytics based on SHAP (SHapley Additive exPlanations) values in Qlik Sense. Building on the previous video where SHAP data was integrated into the prediction app, this session demonstrates how to visualize and interpret feature importance using bar charts, both in terms of directional influence and absolute impact. Learn how to explore model interpretability, uncover key drivers of your predictions, and understand feature effects in detail. Stay tuned as we prepare for more advanced examples in the next video using prediction-specific SHAP data. in this video: 📢 Intro – Introduction by Mike Tarallo Quick greeting and context on the video series. 🔁 Recap – Adding SHAP Data to the App Brief look back at importing SHAP datasets into Qlik Sense. 🧱 Start – Beginning the Analytics Creation Beginning a new sheet and creating visuals manually. 📊 Build – Creating a SHAP Bar Chart Selecting dimensions and measures to analyze feature impact. 🔍 Inspect – Feature Field and SHAP Value Setup Adding and aggregating SHAP values for analysis. 💡 Insight – Understanding Feature Influence Exploring which features have the highest impact on model predictions. ⚖️ Balance – Positive vs Negative Influence Explained Clarifying the directionality of SHAP values and feature effects. 📝 Duplicate – Enhancing with Absolute Value Using fabs() to visualize absolute influence of features. 📈 Magnitude – Why Absolute SHAP Values Matter Understanding the strength of feature impact regardless of direction. ✅ Wrap-Up – Summary and What’s Next Recap and teaser for the next video on prediction-specific SHAP data.