Ablation Study on SCZ, AD, PD, BP, and MG: Model Performance Metrics

Performance metrics from an ablation study on models for five neurological disorders: schizophrenia (SCZ), Alzheimer's disease (AD), Parkinson's disease (PD), bipolar disorder (BP), and myasthenia gravis (MG). The dataset, authored by Jiafang Li and last updated on 2026-05-29, includes evaluation results using AUROC and F1-score on a hold-out test set, with model selection via 5-fold cross-validation.

Use Cases

• Compare model performance across different neurological disorders based on the reported AUROC and F1-scores.
• Validate ablation study methodologies for clinical machine learning models based on the described 5-fold cross-validation process.
• Benchmark new diagnostic algorithms against reported baseline metrics for SCZ, AD, PD, BP, and MG.

Strengths

• Evaluation uses standard metrics (AUROC and F1-score) on a hold-out test set.
• Model selection was conducted via 5-fold cross-validation, a robust methodological step.
• Dataset is openly available under a CC-BY-4.0 license.

Limitations

• Dataset is very small at 10.1 KB, indicating limited scope or summary-level data only.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for certain analyses.

Provenance

Source

figshare

Collection Method

Likely contains results from computational experiments or model evaluations.

Freshness

Last updated 2026-05-29 18:00:10; freshness should be verified.