Identifying Information from Observations with Uncertainty and Novelty

Derek S. Prijatelj, Timothy J. Ireland, Walter J. Scheirer · 2025

Read Identifying Information from Observations with Uncertainty and Novelty by Derek S. Prijatelj, Timothy J. Ireland, Walter J. Scheirer (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-09331.txt.

Description

Prijatelj, Ireland, and Scheirer formalize identifying information from observations in a 2025 paper. A learning machine must handle uncertainty and novelty while keeping its performance. Key questions ask how much information observations give and how much is still needed.

Identifying information is defined as bits that confirm or rule out a hypothesis about what made the data. The authors prove traits of this computation and the sample counts it implies. Cases run from fixed processes to steady random ones, linking finite-step ideas to classic statistics and learning theory. The framework also gives a clean way to spot novel data. Sample needs for one Bayesian learner are pinned down as well.

Catalog record

CREATORDerek S. Prijatelj, Timothy J. Ireland, Walter J. Scheirer
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE129 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2501-09331

Browse the full library · About the Tower of Babel