AI Paper Atlas

Research map / Learning theory and algorithms / Kernels and regression theory

Split-kl and PAC-Bayes-split-kl Inequalities for Ternary Random Variables

Yi-Shan Wu, Yevgeny Seldin

NeurIPS 2022 · Learning theory and algorithms › Kernels and regression theory · approach: PAC-Bayes bounds · 1 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
use split-kl instead of kl inequality for ternary variables as p0 grows toward 1 (symmetric case p1=p-1)split-kl inequality · Simulated ternary variables, n samples, delta=0.05bound minus empirical mean (lower is tighter)Improved
use split-kl instead of Unexpected Bernstein for skewed ternary variable (p1=0.99(1-p0), p-1=0.01(1-p0))split-kl inequality · Simulated skewed ternary variablesbound minus empirical meanImproved
use Empirical Bernstein instead of other bounds on ternary variablesEmpirical Bernstein inequality · Simulated ternary variablesbound tightnessHurt
increase sample size to n=1000 in ternary simulationsEmpirical Bernstein vs Unexpected Bernstein · Simulated ternary variablesbound tightnessMixed
use split-kl on Beta-distributed bounded variables with large variancesplit-kl inequality · Simulated bounded variables, alpha=beta, p=0.5bound tightnessNo clear effect
use informed priors with excess losses and PAC-Bayes-split-kl vs PAC-Bayes-kl and PAC-Bayes-Unexpected-BernsteinPAC-Bayes-split-kl · Linear classification, UCI/LibSVM datasetsgeneralization boundNo clear effect
use CCPBSkl and CCPBUB instead of CCPBB for tandem loss with offsetweighted majority vote bound · Heterogeneous classifier ensembles, UCI datasetsbound valueImproved
use CCPBSkl/CCPBUB instead of TND/CCTND (tandem loss without offset)weighted majority vote bound · Heterogeneous classifier ensembles, UCI datasetsbound valueNo clear effect
apply bounds to reweighting trees in a random forest instead of heterogeneous classifiersCCPBSkl and CCPBUB · Random forest, UCI datasetsbound valueNo clear effect

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy