Research map / Learning theory and algorithms / Sample complexity
The power of absolute discounting: all-dimensional distribution estimation
Moein Falahatgar, Mesrob I Ohannessian, Alon Orlitsky, Venkatadheeraj Pichapati
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 changed | Where | Measured | Effect |
|---|---|---|---|
| use absolute discounting instead of empirical, add-beta, Braess-Sauer, or Paninski estimators | absolute discounting estimator · synthetic power-law alpha=2, k=1000 | expected KL loss | Improved |
| use absolute discounting on piece-wise power-law instead of pure power-law | absolute discounting estimator · piece-wise power-law (1.3, 2, 1.5) | expected KL loss | No clear effect |
| compare Good-Turing + empirical with absolute discounting | Good-Turing + empirical · synthetic power-law | expected KL loss | No clear effect |
| use absolute discounting vs frequency estimator for unseen low-activity cities | absolute discounting estimator · START terrorism forecasting, low-activity cities | forecast of total incidents | Improved |
| use absolute discounting vs frequency estimator for medium-activity cities | absolute discounting estimator · START terrorism forecasting, medium-activity cities | forecast of total incidents | Improved |
| use Braess-Sauer instead of absolute discounting for high-activity cities | Braess-Sauer estimator · START terrorism forecasting, high-activity cities | forecast accuracy | Hurt |
| increase n with fixed k | absolute discounting · power-law alpha=2, k in 100..3000 | expected KL loss | Improved |
Closest papers by meaning
- Model Alignment as Prospect Theoretic Optimization (ICML 2024 · similarity 0.84)
- Learning Parametric Distributions from Samples and Preferences (ICML 2025 · similarity 0.83)
- Tail Competition Explains Scaling in Best-of-N Sampling for Verifiable Problems (NeurIPS 2026 · similarity 0.83)
- RULE: Reinforcement UnLEarning Achieves Forget-retain Pareto Optimality (NeurIPS 2025 · similarity 0.83)
- Multidimensional Bayesian Utility Maximization: Tight Approximations to Welfare (NeurIPS 2025 · similarity 0.83)
- Risk Aware Benchmarking of Large Language Models (ICML 2024 · similarity 0.83)
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.
