Research map / Learning theory and algorithms / Prediction and decision losses
Unbiased Objective Estimation in Predictive Optimization
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (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 changed | Where | Measured | Effect |
|---|---|---|---|
| Parameter perturbation bias correction (Algorithm 2) instead of simple evaluation | predictive price optimization, real-world beer retail dataset · Real-world retail dataset (50 beer products) | estimated gross profit: from 490,502 JPY to 124,477 JPY | Helped |
| Cross-validation style correction (Algorithm 1) corrects optimistic bias but introduces pessimistic underestimation | predictive portfolio optimization · Portfolio simulation (D=50, N=20) | estimate of f(ẑ,θ*) | Mixed |
| Parameter perturbation method instead of CV method | predictive portfolio optimization · Portfolio simulation (D=50, N=20) | bias and variance of estimate of f(ẑ,θ*) | Helped |
| Portfolio resampling heuristic | predictive portfolio optimization · Portfolio simulation (D=50, N=20) | optimistic bias of estimate | No clear effect |
| Hold-out validation with larger N' (hold-out sample size) | predictive portfolio optimization · Portfolio simulation (D=50, N=20) | estimate accuracy | Helped |
| Increasing perturbation parameter h (bias-variance tradeoff) (ablation) | parameter perturbation method · Portfolio simulation, 100 datasets | bias, variance, RMSE | Mixed |
| Parameter perturbation method vs CV with increasing parameter dimensionality d | predictive price optimization, simulated data · Simulation, N=500, d up to 60 | bias relative to f(z*,θ*) | Helped |
| CV method (K=2) in price optimization | predictive price optimization, simulated data · Simulation, N=500, d up to 60 | pessimistic bias | Hurt |
Closest papers by meaning
- The Bias-Variance Tradeoff in Data-Driven Optimization: A Local Misspecification Perspective (NeurIPS 2025 · similarity 0.87)
- On the Optimal Construction of Unbiased Gradient Estimators for Zeroth-Order Optimization (NeurIPS 2025 · similarity 0.86)
- Revisiting Zeroth-Order Optimization: Minimum-Variance Two-Point Estimators and Directionally Aligned Perturbations (ICLR 2025 · similarity 0.85)
- Bayesian Optimisation with Unknown Hyperparameters: Regret Bounds Logarithmically Closer to Optimal (NeurIPS 2024 · similarity 0.84)
- This Too Shall Pass: Removing Stale Observations in Dynamic Bayesian Optimization (NeurIPS 2024 · similarity 0.84)
- ROOT: Rethinking Offline Optimization as Distributional Translation via Probabilistic Bridge (NeurIPS 2025 · similarity 0.84)
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.