Skip to content

Explainability Report

Model explained: logistic_regression

Global SHAP Feature Importance

featuremean_abs_shap
numeric__JobLevel0.549735
numeric__MonthlyIncome0.471733
numeric__TotalWorkingYears0.424355
numeric__YearsSinceLastPromotion0.423696
numeric__NumCompaniesWorked0.405947
categorical__OverTime_No0.359265
numeric__JobSatisfaction0.346377
numeric__EnvironmentSatisfaction0.337649
categorical__OverTime_Yes0.306879
numeric__RelationshipSatisfaction0.293467
numeric__YearsWithCurrManager0.283127
numeric__YearsInCurrentRole0.277629
categorical__BusinessTravel_Travel_Frequently0.253191
categorical__JobRole_Laboratory Technician0.249432
numeric__DistanceFromHome0.231682
numeric__Age0.231313
categorical__MaritalStatus_Single0.230747
categorical__MaritalStatus_Divorced0.203018
numeric__JobInvolvement0.202091
numeric__TrainingTimesLastYear0.196138
categorical__BusinessTravel_Non-Travel0.175278
categorical__JobRole_Sales Representative0.173284
numeric__StockOptionLevel0.171638
numeric__WorkLifeBalance0.154537
numeric__PercentSalaryHike0.152525

Notes

  • SHAP values explain feature contributions to model predictions.
  • Global importance is computed as mean absolute SHAP value.
  • Local explanations are generated for representative validation samples.

Built with VitePress and deployed with GitHub Pages.