Number of the records: 1
Learnability can be undecidable
- 1.0500071 - MÚ 2020 RIV GB eng J - Journal Article
Ben-David, S. - Hrubeš, Pavel - Shay, M. - Shpilka, A. - Yehudayoff, A.
Learnability can be undecidable.
Nature Machine Intelligence. Roč. 1, č. 1 (2019), s. 44-48. E-ISSN 2522-5839
Institutional support: RVO:67985840
Keywords : learnability * machine learning
OECD category: Pure mathematics
Method of publishing: Limited access
http://dx.doi.org/10.1038/s42256-018-0002-3
The mathematical foundations of machine learning play a key role in the development of the field. They improve our understanding and provide tools for designing new learning paradigms. The advantages of mathematics, however, sometimes come with a cost. Gödel and Cohen showed, in a nutshell, that not everything is provable. Here we show that machine learning shares this fate. We describe simple scenarios where learnability cannot be proved nor refuted using the standard axioms of mathematics. Our proof is based on the fact the continuum hypothesis cannot be proved nor refuted. We show that, in some cases, a solution to the ‘estimating the maximum’ problem is equivalent to the continuum hypothesis. The main idea is to prove an equivalence between learnability and compression.
Permanent Link: http://hdl.handle.net/11104/0292219
File Download Size Commentary Version Access Hrubes.pdf 11 1.6 MB Publisher’s postprint require
Number of the records: 1