Machines are given
unlabeled data that is graded after processing.
Once the network learns to perform from training data, it can then be tested against
unlabeled data.
This heuristic training approach holds considerable promise for addressing one of the biggest challenges for neural networks: making correct classifications of previously unknown or
unlabeled data.
With «unsupervised learning,» by contrast, a neural net is shown
unlabeled data and asked simply to look for recurring patterns.
Not exact matches
Since the
data is
unlabeled, the system can't actually deduce the topics of the documents.
The report reveals that the fragrance industry has published safety assessments for only 34 % of the
unlabeled ingredients: «Chemicals range from food additives whose safety in perfumes has not been assessed to chemicals with limited public safety
data, such as synthetic musk fragrances, which accumulate in the human body and may be linked to hormone disruption.»