Hyperparameter Limits for the CB Classifier in Gallstone Prediction

Prosenjit Das published hyperparameter limits for a CatBoost classifier used to predict gallstone disease from tabular data on 2026-06-01. The dataset is associated with a study that employed 38 features and achieved a mean accuracy of 79.58% using 5-fold cross-validation. The research also applied the Sea Lion Optimization Algorithm to select 19 features, improving mean accuracy to 80.42%.

Use Cases

• Tuning CatBoost classifiers for medical prediction tasks based on the described hyperparameter limits.
• Benchmarking optimization algorithms like the Sea Lion Optimization Algorithm (SLOA) for feature selection in tabular data.
• Explaining model predictions for gallstone disease using SHAP, LIME, or DiCE, as referenced in the study.
• Studying the predictive importance of clinical features like C-Reactive Protein (CRP) and Vitamin D for gallstone disease.

Strengths

• Model performance metrics are provided, including a mean accuracy of 79.58% for the base model and 80.42% for the optimized model.
• The study details a specific feature selection process, reducing the feature set from 38 to 19.
• Dataset is openly licensed under CC-BY-4.0.

Limitations

• The dataset is very small at 5.5 KB, indicating limited scope, likely containing only hyperparameter configurations or summary results.
• Row count and specific column-level documentation are absent; data structure must be inferred after download.
• The description focuses on model results; the underlying patient/clinical data used to train the model is not included.

Provenance

Source

Prosenjit Das via figshare

Collection Method

Likely derived from a machine learning study on a public gallstone dataset, employing CatBoost and the Sea Lion Optimization Algorithm.

Freshness

Last updated 2026-06-01 17:51:39; freshness should be verified.