Self-Supervised Learning Algorithm

A self-supervised learning algorithm is a semi-supervised learning algorithm that makes use of a labeling heuristic.

• Context:
  • It can (typically) be used to solve a self-supervised learning task (where a labeling heuristic is provided).
  • …
• Counter-Example(s):
  • a Co-Training Learning Algorithm.
  • a Weakly Supervised Learning Algorithm.
  • a Distant-Supervision Algorithm.
  • a Bootstrapped Learning Algorithm.
• See: Self-Supervised Learning System, Weakly Labeled Dataset.

---

---

References

2008

• (Banko & Etzioni, 2008) ⇒ Michele Banko, and Oren Etzioni. (2008). “The Tradeoffs Between Open and Traditional Relation Extraction.” In: Proceedings of the 46th Annual Meeting of the Association for Computational Linguistics (ACL 2008).
  • QUOTE: As with O-NB, O-CRF’s training process is self-supervised. O-CRF applies a handful of relation-independent heuristics to the PennTreebank and obtains a set of labeled examples in the form of relational tuples. The heuristics were designed to capture dependencies typically obtained via syntactic parsing and semantic role labelling.

2006

• (Mitchell, 2006) ⇒ Tom M. Mitchell. (2006). “The Discipline of Machine Learning." Machine Learning Department technical report CMU-ML-06-108, Carnegie Mellon University.
  • QUOTE: A key research issue here is self-supervised learning and constructing an appropriate graded curriculum.

---