Thumbnail
Access Restriction
Subscribed

Author Pitt, L.
Source ACM Digital Library
Content type Text
Publisher Association for Computing Machinery (ACM)
File Format PDF
Copyright Year ©1989
Language English
Subject Domain (in DDC) Computer science, information & general works ♦ Data processing & computer science
Abstract Inductive inference machines construct programs for total recursive functions given only example values of the functions. $\textit{Probabilistic}$ inductive inference machines are defined, and for various criteria of successful inference, it is asked whether a probabilistic inductive inference machine can infer larger classes of functions if the inference criterion is relaxed to allow inference with probability at least $\textit{p},$ (0 < $\textit{p}$ < 1) as opposed to requiring certainty. For the most basic criteria of success $(\textit{EX}$ and $\textit{BC}),$ it is shown that any class of functions that can be inferred from examples with probability exceeding 1/2 can be inferred deterministically, and that for probabilities $\textit{p}$ ≤ 1/2 there is a discrete hierarchy of inferability parameterized by $\textit{p}.$ The power of probabilistic inference strategies is characterized by equating the classes of probabilistically inferable functions with those classes that can be inferred by $\textit{teams}$ of inductive inference machines (a parallel model of inference), or by a third model called $\textit{frequency}$ inference.
ISSN 00045411
Age Range 18 to 22 years ♦ above 22 year
Educational Use Research
Education Level UG and PG
Learning Resource Type Article
Publisher Date 1989-04-01
Publisher Place New York
e-ISSN 1557735X
Journal Journal of the ACM (JACM)
Volume Number 36
Issue Number 2
Page Count 51
Starting Page 383
Ending Page 433


Open content in new tab

   Open content in new tab
Source: ACM Digital Library