Asked about Finding and fixing bad data
You suspect many labels in your training set are wrong. How do you find them?
An answer that works
Get out-of-fold predictions for every example, rank by how confidently the model disagrees with the given label, and review the top of that list by hand. Estimate the error rate from the reviewed sample, fix what you can, and fix the labelling process that produced them.
Then they push further
Is it safe to just drop every example the model disagrees with?
No. The model may be wrong, and the hard, correct examples are exactly what it disagrees with. Dropping them teaches it only the easy cases. Review by hand, at least on a sample.
What gets marked up
Marked up for out-of-fold predictions, human review, and fixing the source of errors rather than only the data.
Reading an answer is not being able to give one. The section that teaches this is Finding and fixing bad data, 4 hours of Data quality & labelling.
Learn this properlyAlso asked about Data quality & labelling