Tosca Lechner

Impossibility of Characterizing Distribution Learning -- a simple solution to a long-standing problem

2023 · arXiv (Cornell University)

We consider the long-standing question of finding a parameter of a class of probability distributions that characterizes its PAC learnability. We provide a rather surprising answer - no such parameter exists. Our techniques allow us …

On the Computability of Robust PAC Learning

2024 · arXiv (Cornell University)

We initiate the study of computability requirements for adversarially robust learning. Adversarially robust PAC-type learnability is by now an established field of research. However, the effects of computability requirements in PAC-type frameworks are only just …

On the Computability of Multiclass PAC Learning

2025 · arXiv (Cornell University)

We study the problem of computable multiclass learnability within the Probably Approximately Correct (PAC) learning framework of Valiant (1984). In the recently introduced computable PAC (CPAC) learning framework of Agarwal et al. (2020), both learners …