Editing Algorithmic learning theory (section)

==Other identification criteria==

Learning theorists have investigated other learning criteria,<ref>Jain, S. et al (1999): ''Systems That Learn'', 2nd ed. Cambridge, MA: MIT Press.</ref> such as the following.

* ''Efficiency'': minimizing the number of data points required before convergence to a correct hypothesis.
* ''Mind Changes'': minimizing the number of hypothesis changes that occur before convergence.<ref>Luo, W. & Schulte, O. (2005), ''[http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.72.7998&rep=rep1&type=pdf Mind Change Efficient Learning]'', in Peter Auer & Ron Meir, ed., Proceedings of the Conference on Learning Theory (COLT), pp. 398-412</ref>

Mind change bounds are closely related to [[Winnow algorithm#Mistake bounds|mistake bounds]] that are studied in [[statistical learning theory]].<ref>{{cite journal |last1=Jain |first1=Sanjay |last2=Sharma |first2=Arun |title=On a Generalized Notion of Mistake Bounds |journal=Information and Computation |date=May 2001 |volume=166 |issue=2 |pages=156–166 |doi=10.1006/inco.2000.3001 }}</ref> Kevin Kelly has suggested that minimizing mind changes is closely related to choosing maximally simple hypotheses in the sense of [[Occam’s Razor]].<ref>{{cite journal |last1=Kelly |first1=Kevin T. |title=Ockham's razor, empirical complexity, and truth-finding efficiency |journal=Theoretical Computer Science |date=September 2007 |volume=383 |issue=2–3 |pages=270–289 |doi=10.1016/j.tcs.2007.04.009 |url=https://figshare.com/articles/journal_contribution/6492356 }}</ref>