Great work on isotree. May I ask you a question? Is there a way to tell which column or columns contributes the most to the high anomaly score, in high dimensional data? It would be tremendously helpful in medical fields.
There's no such functionality in this package but you can use black-box prediction explainers from other libraries based on e.g. SHAP or local sensitivity.
Great work on isotree. May I ask you a question? Is there a way to tell which column or columns contributes the most to the high anomaly score, in high dimensional data? It would be tremendously helpful in medical fields.