Model Performance for Rice Grain Width Prediction Using Five-Fold Cross-Validation

Evaluation metrics for six statistical models predicting rice grain width. The 5.5 KB XLS file, authored by Rupam Basu and last updated in March 2026, compares Spike-and-Slab, Bayesian LASSO, BSLMM, Ridge, LASSO, and OLS models. Performance is assessed using RMSE, MAE, and Predictive Coverage from a five-fold cross-validation.

Use Cases

• Compare the predictive accuracy of Bayesian vs. frequentist models based on reported RMSE and MAE metrics.
• Assess model reliability for genomic selection in breeding based on the predictive coverage statistic.
• Benchmark the Spike-and-Slab prior model against other variable selection methods like LASSO and Ridge for GWAS.
• Inform methodological choices for predicting grain width (GRWD) in rice based on a five-fold cross-validation framework.

Strengths

• Directly compares six distinct modeling approaches (Spike-and-Slab, Bayesian LASSO, BSLMM, Ridge, LASSO, OLS).
• Evaluates models using three standard metrics (RMSE, MAE, Predictive Coverage).
• Uses a five-fold cross-validation methodology, which suggests a structured evaluation process.
• Released under a permissive CC-BY-4.0 license for reuse.

Limitations

• The dataset is very small at 5.5 KB, indicating limited scope.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment for large-scale analysis.

Provenance

Source

figshare, author Rupam Basu.

Collection Method

Likely contains results from a computational experiment evaluating models for a genome-wide association study (GWAS) on rice.

Freshness

Last updated 2026-03-17 17:51:08; freshness should be verified.