Skip to main content

Table 11 XGBoost accuracies (mean ± standard deviation in %) for the external AIBL data set

From: Data analysis with Shapley values for automatic subject selection in Alzheimer’s disease data sets using interpretable machine learning

Exclusion method Number of training subjects excluded
(base model) 0 50 100 150 200 250
Random (-) 60.00 ±5.00 55.00 ±3.98 57.50 ±4.06 55.71 ±2.37 52.14 ±5.35 51.07 ±4.80
LOO (LR) 60.00 ±5.00 58.93 ±3.29 61.79 ±4.53 63.21 ±3.93 58.21 ±3.93 59.64 ±4.80
LOO (RF) 60.00 ±5.00 59.64 ±5.99 61.43 ±5.71 61.79 ±2.79 61.07 ±4.64 56.07 ±4.53
Data Shapley (LR) 60.00 ±5.00 61.07 ±3.73 61.43 ±3.85 62.14 ±3.98 63.21 ±5.31 61.07 ±1.92
Data Shapley (RF) 60.00 ±5.00 64.64 ±5.64 62.86 ±2.86 62.86 ±1.75 61.79 ±2.29 58.57 ±2.37
  1. Different methods were used to identify and focus on the training subjects with the most informative data. Ten repetitions with different seeds were performed for every exclusion data set. The best results are highlighted in bold