Skip to main content
Fig. 4 | BMC Medical Research Methodology

Fig. 4

From: Don’t dismiss logistic regression: the case for sensible extraction of interactions in the era of machine learning

Fig. 4

a-c Diabetes SHAP summary plots for: a) Logistic Regression, b) Random Forest, and c) Hybrid Approach; d) Odds-ratios of predictors derived using logistic regression prior to application of hybrid procedure; e) Odds-ratios of predictors derived from hybrid approach logistic regression coefficients; f) SHAP interaction scores for top 10 ranked interactions

Back to article page