DiCE Counterfactual Explanations for Gallstone Status Prediction

Five counterfactual explanations generated using the DiCE method for a specific instance (index 116) of the SLOA_CB model. The data demonstrates how variations in CRP (F36) and Vitamin D (F38) levels influence the model's prediction for Gallstone Status. The 5.5 KB XLS file was authored by Prosenjit Das and last updated on June 1, 2026.

Use Cases

• Evaluating the sensitivity of a gallstone prediction model to changes in CRP and Vitamin D levels based on the described counterfactuals.
• Benchmarking the DiCE (Diverse Counterfactual Explanations) method's output for a single clinical instance.
• Studying feature influence in a binary classification task for gallstone status.

Strengths

• Provides a concrete, reproducible example with five specific counterfactual instances.
• File size is small (5.5 KB), ensuring quick download and inspection.
• Released under a permissive CC-BY-4.0 license, allowing for broad reuse.

Limitations

• The dataset is extremely small, containing only five rows of counterfactual data for a single model instance.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count for the original model training data is unknown, which limits context for the explanations.

Provenance

Source

Prosenjit Das via figshare

Collection Method

Generated using the DiCE (Diverse Counterfactual Explanations) method applied to the SLOA_CB model.

Freshness

Last updated 2026-06-01 17:51:55; freshness should be verified.