Interpretion guidelines for classification scores
#2
by
jakubuilds
- opened
Thank you so much for sharing this resource with the community! I'm using this model to help identify racist tweets for a study on confronting racism in social media, but I was curious to know more about how to interpret the score_xnumbers that are being generated along with each classification label. Any additional information would be greatly appreciated. Thanks!
Thanks for your interest in our model! These scores are confidence probability scores produced by the model. They represent how confident the model is when it predicts that label. The scores are normalized between 0 and 1, but if you multiply them by 100, you can then interpret the percentage of prediction confidence.
Thank you!
jakubuilds
changed discussion status to
closed