Site Loader
Rock Street, San Francisco

extbf{Background and Objective: }
One of the most important hospital quality indicators is hospital readmission rate. Because unplanned readmissions have high human and financial consequences, many predictive models that try to make early predictions of readmission risk emerged. However, most of developed models are trying to predict readmission risk as a single label, binary problem (patient will be readmitted or not). Such models do not answer the question extit{“why is a patient at risk for readmission”}. Our objective is to develop interpretable predictive models that can explain why patient is likely to be readmitted (with which diagnoses) which will help making shift from predictive to prescriptive exploitation of predictive models in real medical practice.

In order to address this problem we propose a method that integrates data-driven and expert-driven hierarchy for multi-label and hierarchical multi-label classification. Our method is based on well-known algorithm called Predictive Clustering Trees. In this case, we predict diagnoses and symptoms that are likely to appear on readmitting for specific patient. For the hierarchical multi-label classification task we inspect and interpret the influence of data-driven and expert-driven hierarchies on predictive performance. Experiments are conducted on pediatric population of patients from state of California.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!

order now

Predictive models performance is thoroughly evaluated using example-based, label-based and ranking-based performance measures. Results showed that usage of structure provided by domain knowledge hierarchy improved predictive accuracy for approximately 40\%. Drastic improvement in performance is visible in other measures such as precision, recall, ranking loss etc. Additionally, results are interpreted and discussed by medical doctors.

It is shown that integration of structure (both data- and expert-driven) increases predictive performance compared to flat models and that multi-label classification models based on Predictive Clustering Trees allow building of accurate and highly interpretable solutions. Additionally, models that are built allowed interpretation of the results by analysis of groups obtained from predictive clustering trees.

Post Author: admin


I'm Anna!

Would you like to get a custom essay? How about receiving a customized one?

Check it out