Scenario: A small bank wants to improve its loan approval process by predicting loan default risk. They have collected data on applicants' credit scores, annual incomes, loan amounts, and employment status. The bank is considering using either SVM or decision trees for this task and seeks advice on which model to choose based on ease of use, interpretability, and predictive performance. Question 6: Considering the bank's needs for interpretability and accurate risk prediction, would you recommend using SVM or decision trees? Outline the key factors influencing your decision and suggest a simple method for evaluating the chosen model's performance. Hints to nudge you towards evaluating key aspects of each model's suitability for the bank's scenario: 1. Clarity for Non-Experts: Which model provides a straightforward way to see and explain decisions? 2. Handling Non-Linearity: Is the bank's data simple or complex? One model is better for complex patterns. 3. Preprocessing Needs: Think about the effort needed to prepare the data for each model. 4. Evaluating Success: Beyond accuracy, what metrics could tell us more about the model's performance