logical-accuracy

name	logical-accuracy
description	Detect logical outliers in clinical variables using a multi-model ensemble (Quantile Regression + Gradient Boosting + Autoencoder for numeric; OneClassSVM + IsolationForest + Autoencoder for categorical). Uses Claude CLI (no API key required). Use for LYDUS data quality assessment of logical consistency in event, diagnosis, prescription, and procedure variables.
tier	community
category	lydus
parameters	{"quiq_path":{"description":"Path to QUIQ-format CSV file (output of quiq skill).","type":"string"},"save_path":{"description":"Directory path to save output files (logical_accuracy_summary.csv, logical_accuracy_total.txt, outlier_*.csv).","type":"string"},"operation_type_manual":{"description":"True = manually specify target_variable; False = automatic top-N by count.","type":"boolean"},"target_variable":{"description":"Target variable name (used when operation_type_manual=True).","type":"string"},"automatic_num":{"description":"Number of top variables to analyze automatically (used when operation_type_manual=False). Default 5.","type":"integer"},"recommend_num":{"description":"Number of correlated variables recommended by LLM. Default 5.","type":"integer"}}

Logical Accuracy

Detects logical outliers in clinical variables using a multi-model anomaly detection ensemble. A value is flagged as an outlier only when agree — minimizing false positives.