UPDF AI

Some contributions to fixed-distribution learning theory

M. Vidyasagar,S. Kulkarni

2000 · DOI: 10.1109/9.839945
IEEE Transactions on Automatic Control · 3 Citations

TLDR

Two new notions of learnability are introduced; these are probably uniformly approximately correct (PUAC) learnability which is a stronger requirement than the widely studied PAC learnability, and minimal empirical risk (MER) learnable, which is an stronger requirementthan the previously defined notions of "solid" or "potential" learnability.