Unlabeled Training Dataset

An Unlabeled Training Dataset is a training dataset with Unlabeled Training Records.

• Example(s):
  • an Unannotated Corpus.
  • …
• Counter-Example(s):
  • a Labeled Training Record Set.
  • an Unlabeled Testing Dataset.
• See: Semi-Supervised Learning Algorithm.

---

---

References

2000

• (Nigam et al., 2000) ⇒ Kamal Nigam, Andrew McCallum, Sebastian Thrun, and Tom M. Mitchell. (2000). “Text Classification from Labeled and Unlabeled Documents Using EM.” In: Machine Learning, 39(2/3). doi:10.1023/A:1007692713085
  • QUOTE: ... over the large set of unlabeled training data can provide useful information to construct a more accurate classifier that considers ...

    We now proceed to the main topic of this paper: how unlabeled data It can be used to improve a text classifier.

    In this section, we provide empirical evidence that combining labeled and unlabeled training documents using EM outperforms traditional naive Bayes, which trains on labeled documents alone. ...

---