Layerwise computability and image randomness (Q1694009)

Any reasonable notion of randomness should obey the property that the image \(F(x)\) of an object \(x\) is random with respect to the image distribution \(F(P)\) if and only if it has a \(P\)-random \(F\)-preimage. For computable distributions \(P\) and mappings \(F\) and Martin-Löf randomness this result is known for a long time; for layerwise computable mappings it was mentioned in [the second author and \textit{C. Rojas}, Lect. Notes Comput. Sci. 5555, 549--561 (2009; Zbl 1248.03066)]. The paper provides a proof of this result and discusses quantitative results on randomness deficiency and an application to random closed sets and many-valued mappings. To this end in Section 2 some results on layerwise computability are derived. Section 3 discusses image randomness, and the last section contains a quantitative version of the theorem on image randomness from Section 3, a quantitative version for expectation-bounded deficiency, an application to random closed sets and considerations about non-deterministic mappings.