AI Paper Atlas

Research map / Learning theory and algorithms / Prediction and decision losses

Advice Querying under Budget Constraint for Online Algorithms

Ziyad Benomar, Vianney Perchet

NeurIPS 2023 · Learning theory and algorithms › Prediction and decision losses · approach: Algorithms with explicit predictors · 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
query the prediction after renting t days instead of at t=0 (deterministic Algorithm 1)Ski-rental with time-dependent oracle accuracy · Ski rental, b=50,100, pt=0.95-0.4exp(-t/5)competitive ratioImproved
query the prediction after renting t days (randomized Algorithm 2) with b=50Ski-rental randomized algorithm · Ski rental, b=50competitive ratioNo clear effect
use randomized instead of deterministic algorithm after the binary predictionSki-rental with accuracy-p oracle · Ski rental, p in (0.5,1)competitive ratio boundImproved
use randomized instead of deterministic algorithm when oracle accuracy p is very close to 1Ski-rental with accuracy-p oracle · Ski rental, p near 1competitive ratio bound: 1 + 1/b instead of 1Hurt
tune lambda optimally as a function of p instead of fixed lambdaAlgorithm 1 deterministic ski rental · Ski rental, p>=0.5competitive ratio bound: 1 + 2sqrt(p(1-p)) <= 2Improved
compare experimental to theoretical competitive ratio of deterministic Algorithm 1Algorithm 1 · Ski rental b=50, x~U[1,4b]competitive ratioImproved
increase prediction budget B in ADATHRESHADATHRESH secretary algorithm · Secretary, N=1000, B<=20success probabilityImproved
add memory of the best observed value after restartADATHRESH · Secretary, N=1000success probabilityImproved
increase oracle accuracy p toward 1 in imperfect-prediction ADATHRESHADATHRESH with noisy oracle · Secretary, B<=20success probability lower bound qB(p): limit 1 - 1/(1 + W(p/(e(1-p))))Improved
trust oracle only if p above about 0.73, else use 1/e-ruleADATHRESH with noisy oracle · Secretarysuccess probability: qB(p)=1/e for p <= 0.73Improved
query job sizes of B jobs only after waiting or running RR firstB-clairvoyant scheduling · Non-clairvoyant schedulingcompetitive ratio: 2No clear effect
give sublinear B=o(N) job-size queriesPreemptive non-clairvoyant scheduling · Scheduling lower boundcompetitive ratio: 2No 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