Pith. sign in

REVIEW 2 cited by

Second-Order Asymptotically Optimal Statistical Classification

Not yet reviewed by Pith; the record is open.

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 1806.00739 v3 pith:XSNYVDIR submitted 2018-06-03 cs.IT cs.LGmath.IT

Second-Order Asymptotically Optimal Statistical Classification

classification cs.IT cs.LGmath.IT
keywords distributionsclassificationerrorproblemaccordingbinarygeneratedprobability
verification ladder T0 review T1 audit T2 compute T3 formal T4 reserved
0 comments
Share X Bluesky LinkedIn Reddit HN
read the original abstract

Motivated by real-world machine learning applications, we analyze approximations to the non-asymptotic fundamental limits of statistical classification. In the binary version of this problem, given two training sequences generated according to two {\em unknown} distributions $P_1$ and $P_2$, one is tasked to classify a test sequence which is known to be generated according to either $P_1$ or $P_2$. This problem can be thought of as an analogue of the binary hypothesis testing problem but in the present setting, the generating distributions are unknown. Due to finite sample considerations, we consider the second-order asymptotics (or dispersion-type) tradeoff between type-I and type-II error probabilities for tests which ensure that (i) the type-I error probability for {\em all} pairs of distributions decays exponentially fast and (ii) the type-II error probability for a {\em particular} pair of distributions is non-vanishing. We generalize our results to classification of multiple hypotheses with the rejection option.

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Forward citations

Cited by 2 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score.

  1. Black-Box Detection of LLM-Generated Text Using Generalized Jensen-Shannon Divergence

    cs.LG 2025-10 conditional novelty 6.0

    A reference-based detector that scores text by the generalized Jensen–Shannon gap between its surprisal-state transition matrix and fixed human/machine references.

  2. Black-Box Detection of LLM-Generated Text Using Generalized Jensen-Shannon Divergence

    cs.LG 2025-10 unverdicted novelty 5.0

    SurpMark detects machine-generated text by estimating state-transition matrices from discretized surprisals and scoring them with generalized Jensen-Shannon divergence to human versus machine references.