Your data science team is developing a credit-risk prediction model at a financial services company. After testing logistic regression, decision trees, and support vector machines individually, none achieves the desired accuracy or robustness. You want to combine these models to leverage their complementary strengths and reduce weaknesses. Which approach is MOST likely to improve the model’s performance?