Model Comparison Statistical Significance Results

Excel tables contain paired statistical significance analysis comparing a MaxGRNet model against four baseline models using 5-fold cross-validation. Md Mehedi Hasan Santo published the results on figshare in April 2026. The 5.5 KB file includes mean fold-wise differences, confidence intervals, and corrected p-values.

Use Cases

• Compare mean fold-wise differences between MaxGRNet and baseline models to assess performance superiority.
• Interpret Holm-Bonferroni corrected p-values to determine statistically significant differences across multiple baseline comparisons.
• Analyze Wilcoxon signed-rank test p-values for paired, non-parametric significance testing.
• Use 95% confidence intervals computed from paired fold-wise differences to quantify uncertainty in model comparison.

Strengths

• Results are derived from a structured 5-fold cross-validation procedure.
• Analysis includes multiple statistical tests: paired t-test, Holm-Bonferroni correction, and Wilcoxon signed-rank test.
• Data is openly licensed under CC-BY-4.0.

Limitations

• The dataset is very small, only 5.5 KB, limiting the scope of analysis.
• The underlying dataset DS1 and the specific baseline models are not described.
• The number of rows and columns is unknown.

Provenance

Source

figshare

Collection Method

Statistical analysis of machine learning model performance on dataset DS1.

Freshness

Last updated in April 2026.