Hypergraph Neural Network Performance Comparison for Seizure Prediction

Comparison data evaluates HP-GNN model performance across three datasets, including results on 22 pediatric patients. The dataset was authored by Masoud Amiri and last updated in April 2026. It contains evaluation metrics in a 5.5 KB Excel file.

Use Cases

• Benchmark HP-GNN model accuracy against a reported 84% seizure detection accuracy metric.
• Analyze performance variation across datasets involving 22 pediatric patients and new patients.
• Compare model results using features derived from 256-dimensional data representations.
• Evaluate the impact of hypergraph convolution layers and prediction heads on task performance.

Strengths

• Direct model performance comparison across three distinct datasets.
• Includes specific result metrics such as an achieved 84% detection accuracy.
• Published under a permissive CC-BY-4.0 license for reuse.

Limitations

• Extremely small data volume at only 5.5 KB, indicating summary results only.
• Lacks raw input data, containing only aggregated performance metrics.
• Unknown row count and specific column structure limits detailed analysis.

Provenance

Source

figshare, authored by Masoud Amiri.

Collection Method

Presumably derived from experimental evaluation of HP-GNN models on seizure prediction tasks.

Freshness

Last updated April 2026.