data/final_report.json

{
  "cross_validation": {
    "n_folds": 1,
    "metrics": {
      "test_auc": {
        "mean": 0.8927442022563707,
        "std": 0.0
      },
      "test_auprc": {
        "mean": 0.6400066633536523,
        "std": 0.0
      },
      "test_f1": {
        "mean": 0.5650997293775684,
        "std": 0.0
      },
      "test_mcc": {
        "mean": 0.4626629343517019,
        "std": 0.0
      },
      "test_kappa": {
        "mean": 0.3640189807009344,
        "std": 0.0
      },
      "test_brier": {
        "mean": 0.25321514268945916,
        "std": 0.0
      },
      "test_sensitivity": {
        "mean": 0.9778009018349713,
        "std": 0.0
      },
      "test_specificity": {
        "mean": 0.5760610868697182,
        "std": 0.0
      }
    }
  },
  "calibration": {
    "temperature": 1.6994811296463013,
    "ece_before": 0.5400210692875276,
    "ece_after": 0.4407205309405988
  },
  "best_fold": 0,
  "datasets": {
    "oasis_slices": 86437,
    "ixi_slices": 34762
  }
}