SHAP waterfall plot. by Wentao Yang (205781)

“…Across 10 models, CatBoost performed best on the test set (AUC = 0.970, accuracy = 0.920, F1 = 0.918), with robust calibration and decision-curve net benefit. SHAP interpretation ranked eGDR among the most influential predictors: SHAP summary and dependence plots indicated that higher eGDR decreased the model’s predicted probability of frailty. …”

SHAP decision plot. by Wentao Yang (205781)

“…Across 10 models, CatBoost performed best on the test set (AUC = 0.970, accuracy = 0.920, F1 = 0.918), with robust calibration and decision-curve net benefit. SHAP interpretation ranked eGDR among the most influential predictors: SHAP summary and dependence plots indicated that higher eGDR decreased the model’s predicted probability of frailty. …”

SHAP dependence plots. by Wentao Yang (205781)

“…Across 10 models, CatBoost performed best on the test set (AUC = 0.970, accuracy = 0.920, F1 = 0.918), with robust calibration and decision-curve net benefit. SHAP interpretation ranked eGDR among the most influential predictors: SHAP summary and dependence plots indicated that higher eGDR decreased the model’s predicted probability of frailty. …”

SHAP dependence plots with interaction coloring. by Wentao Yang (205781)

“…Across 10 models, CatBoost performed best on the test set (AUC = 0.970, accuracy = 0.920, F1 = 0.918), with robust calibration and decision-curve net benefit. SHAP interpretation ranked eGDR among the most influential predictors: SHAP summary and dependence plots indicated that higher eGDR decreased the model’s predicted probability of frailty. …”

Tissue, days post-infection (dpi) and the top 10 most significant genes with increased and decreased expression with valid gene symbols for the response contrasts. by Gillian P. McHugo (8965919)

Median SHAP values (calculated for CSCHF at age 5000 days) for binary features in the competing risks RSF model for different disease categories. The features are sorted in decreas... by Eric Wang (30587)

SHAP values for the top 15 factors implicated in the model. by Erin L. Funnell (20189891)

Shapley Additive Explanations (SHAP) summary plot. by Yue Lv (549862)

SHAP Tree explainer for (a) Model 1: Presentation and (b) Model 6: Closures. by Nathan C. Hurley (10510304)

GI-1d-a: Comparison of estimates of people for dataset A (left bar) and B (right bar) per region. The largest discrepancies occur in areas of NE, SE and SW Germany and adjacent reg... by Isabell Schmidt (6331013)

Innovation, nightlights and a continuous measure of city size. by Saul Estrin (8629173)

Selection frequencies plotted in decreasing order. by Hannah P. Rowe (7830599)

Trust change by gender. by Alain Abi-Rizk (8549358)

“…Results indicate that respondnatds from North America and Europe Experienced an increase in trust where as Africa and South America Witnessed a decline in trust. Males exhibited a greater decline in trust than females (p = 0.038), and undergraduate degree holders showed the largest decrease (p = 0.001). …”

Trust change by age group. by Alain Abi-Rizk (8549358)

“…Results indicate that respondnatds from North America and Europe Experienced an increase in trust where as Africa and South America Witnessed a decline in trust. Males exhibited a greater decline in trust than females (p = 0.038), and undergraduate degree holders showed the largest decrease (p = 0.001). …”

Trust change by education level. by Alain Abi-Rizk (8549358)

“…Results indicate that respondnatds from North America and Europe Experienced an increase in trust where as Africa and South America Witnessed a decline in trust. Males exhibited a greater decline in trust than females (p = 0.038), and undergraduate degree holders showed the largest decrease (p = 0.001). …”

Trust change by source of science information. by Alain Abi-Rizk (8549358)

“…Results indicate that respondnatds from North America and Europe Experienced an increase in trust where as Africa and South America Witnessed a decline in trust. Males exhibited a greater decline in trust than females (p = 0.038), and undergraduate degree holders showed the largest decrease (p = 0.001). …”

Trust levels before and after the COVID-19 pandemic (n = 10,000). by Alain Abi-Rizk (8549358)

Distribution of individual change in trust in science before and after the pandemic (n = 10,000). by Alain Abi-Rizk (8549358)

Feature importance and beeswarm plot determined by mean absolute SHAP values for the DeepSurv model. by Theophilus Gyedu Baidoo (21162511)

Ethnic identity as a moderator among APIDA students. by Ronna Bañada (19233063)