Travel Insurance Claim Propensity Datathon

Logistic regression model - scaled numerical features with SMOTE applied (80:20). High variance / low co-relation features eliminated [Duration, Commission] - Datathon Submission only