You deserve a better ROC curve

<p>All source in this article is available&nbsp;<a href="https://github.com/peterbbryan/rainbow-roc/blob/main/scripts/sample_use.py" rel="noopener ugc nofollow" target="_blank">on Github</a>! Feel free to reuse graphics in any context without attribution (no rights reserved).</p> <h1>Motivation</h1> <p>Receiver operating characteristic (ROC) curves are invaluable for understanding the behavior of a model. There is more information packed into those curves than you typically see, though. By visualizing the threshold value, we can learn more about our classifier. As a bonus, retaining this information makes ROC curves more intuitive.</p> <p><a href="https://towardsdatascience.com/you-deserve-a-better-roc-curve-970617528ce8"><strong>Learn More</strong></a></p>