A hybrid approach for modeling bicycle crash frequencies: Integrating random forest based SHAP model with random parameter negative binomial regression model

Computer Science

• Random Decision Forest 100%
• Hybrid Approach 100%
• Shapley Additive Explanation 50%
• Extreme Gradient Boosting 33%
• Explanatory Power 33%
• Information Criterion 33%
• Prediction Accuracy 16%
• Interpretability 16%
• Machine Learning Algorithm 16%
• Root Mean Squared Error 16%
• Mean Absolute Error 16%
• Gradient Boosting 16%
• Traffic Analysis 16%
• Support Vector Machine 16%
• Model Prediction 16%
• Interpretable Machine Learning 16%

Engineering

• Random Forest 100%
• Hybrid Approach 100%
• Model Prediction 16%
• Policymakers 16%
• Root-Mean-Squared Error 16%
• Machine Learning Algorithm 16%
• Mean Absolute Error 16%
• Akaike Information Criterion 16%
• Learning System 16%
• Support Vector Machine 16%
• Mathematical Model 16%

Earth and Planetary Sciences

• Machine Learning 100%
• London 50%
• Support Vector Machine 50%