Deep learning in insurance: Accuracy and model interpretability using TabNet

Computer Science

• Interpretability 100%
• Extreme Gradient Boosting 100%
• Deep Learning Method 100%
• Deep Learning Model 66%
• Sparsity 33%
• Logistic Regression 33%
• Multiple Domain 33%
• Dominant Model 33%
• Logistic Regression Model 33%
• Machine Learning 33%
• Learning System 33%
• Leaning Parameter 33%

Mathematics

• Interpretability 100%
• Deep Learning Method 100%
• Generalized Linear Model 50%
• Logistic Regression 25%
• Running Time 25%
• Logistic Regression Model 25%

Economics, Econometrics and Finance

• Pricing 100%
• Logit Model 66%
• Industry 33%
• Machine Learning 33%