Kokiri: Random Forest-Based Cohort Comparison and Characterization



 View presentation:2022-10-17T15:55:00ZGMT-0600Change your timezone on the schedule page
2022-10-17T15:55:00Z

Exemplar figure, described by caption below — The figure shows Kokiri integrated in Coral, comparing six lung cancer patient cohorts of different race and gender. A ranked list shows the most important attributes to differentiate the cohorts. The overall separability is shown with stacked bar charts. A scatterplot gives an overview of the predicted cohort affiliations of the patients, and a second ranked list displays all items, the cohorts they belong to, and the probabilities to belong to any of the cohorts based on the data.

The live footage of the talk, including the Q&A, can be viewed on the session page, BioMedical AI: Session 2 and Panel.

Abstract