Comparison of human consensus ratings and M.E.D.V.I.S. algorithm classifications for figure quality.
<p>(a) Confusion matrix showing agreement between human majority votes (<i>Good</i> vs. <i>Bad</i>) and algorithm predictions. (b) Per-figure lollipop plot of the proportion of human raters voting <i>Good</i> (horizontal axis), with points colored green when...
محفوظ في:
| المؤلف الرئيسي: | |
|---|---|
| مؤلفون آخرون: | , , , , , |
| منشور في: |
2025
|
| الموضوعات: | |
| الوسوم: |
إضافة وسم
لا توجد وسوم, كن أول من يضع وسما على هذه التسجيلة!
|
| الملخص: | <p>(a) Confusion matrix showing agreement between human majority votes (<i>Good</i> vs. <i>Bad</i>) and algorithm predictions. (b) Per-figure lollipop plot of the proportion of human raters voting <i>Good</i> (horizontal axis), with points colored green when the algorithm agreed and red when it disagreed. The shaded region indicates the decision threshold (≥0.5 = <i>Good</i>), and right-side annotations show the algorithm’s assigned label for each figure.</p> <p>(TIF)</p> |
|---|