Fig. 1: Individual factor importance for predicting current creativity change.
From: Neuropsychological insights into creativity in people with Parkinson’s disease
Plot displays SHAP values (color code: low = blue, red = high, and showing non-linearity of associations). Note. This SHAP beeswarm plot visually summarizes the influence of predictors on a model’s predictions, illustrating both the magnitude and direction of their contributions. Each dot represents an individual data point, with the x-axis indicating the SHAP value (the predictors contribution to the prediction) and the color denoting the predictor’s value (e.g., blue for low values, red for high values). For the predictor post diagnosis: creativity change, low values (e.g., reduced post-diagnosis: creativity change) correspond to negative SHAP values, indicating a suppressive effect on the prediction of current creativity change. Conversely, high values (e.g., increased post-diagnosis: creativity change) are associated with positive SHAP values, signifying a promotive effect on the prediction.
