Correction of noisy labels via mutual consistency check

  title={Correction of noisy labels via mutual consistency check},
  author={Sahely Bhadra and Matthias Hein},
Label noise can have severe negative effects on the performance of a classifier. Such noise can either arise by adversarial manipulation of the training data or from unskilled annotators frequently encountered in crowd sourcing (e.g. Amazon mechanical turk). Based on the assumption that an expert has provided some fraction of the training data, where labels can be assumed to be true, we propose a new pre-processing method to identify and correct noisy labels via a mutual consistency check using… CONTINUE READING