1 paper
Aleksandr Karakulev, Dave Zachariah, Prashant Singh
We present an adaptive approach for robust learning from corrupted training sets. We identify corrupted and non-corrupted samples with latent Bernoulli variables and thus formulate…