Get: Outliers Detection
Identifies extreme values using methods such as Z‑Score, IQR, Isolation Forest, or DBSCAN.
Endpoint:
/eda/outliers_detection
Input Parameters:
- usecaseVersion (integer, default: 1)
- feature_col (string, required)
- method (string, required: IQR, Z‑Score, Isolation Forest, DBSCAN)
- plot_type (string, required: boxplot or violin)
- z_threshold (number, optional; for Z‑Score method)
- eps (number, optional; for DBSCAN method)
-
usecaseId (string, required)
Output:
JSON with flagged outliers, summary statistics, and visual plots.