Research map / Learning theory and algorithms / Sample complexity
Fundamental Tradeoffs in Learning with Prior Information
Anirudha Majumdar
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 |
|---|---|---|---|
| Weight risk by the learner's prior and use the resulting prioritized-risk framework | Learning problems where reality may not conform to the learner's prior · General learning problems | prioritized risk lower bound | Improved |
| Reduce estimation to testing with a prior-weighted test and a (δ, π)-packing | Estimation problems · General estimation problems | prioritized risk lower bound | Improved |
| Apply the generalized Fano inequality using the Donsker–Varadhan change-of-measure inequality | General learning problems with unbounded losses · General learning problems | Bayes risk lower bound | Improved |
| Use a more concentrated Beta(α=1, β=4) prior | Bayesian inference for Bernoulli mean estimation · Bernoulli mean estimation | learner-specific prioritized risk | Improved |
| Use regressors with smaller L2 norm | Logistic regression with directional priors · Logistic regression | prioritized risk lower bound | Improved |
| Restrict the available policy set from 50 policies to 5 | RL agent in Zipfian Gridworld environments · Zipfian Gridworld environments | prioritized risk lower bound | Hurt |
Closest papers by meaning
- Assouad, Fano, and Le Cam with Interaction: A Unifying Lower Bound Framework and Characterization for Bandit Learnability (NeurIPS 2024 · similarity 0.89)
- Causal Bandits: The Pareto Optimal Frontier of Adaptivity, a Reduction to Linear Bandits, and Limitations around Unknown Marginals (ICML 2024 · similarity 0.87)
- Logarithmic Smoothing for Pessimistic Off-Policy Evaluation, Selection and Learning (NeurIPS 2024 · similarity 0.87)
- Information Complexity of Stochastic Convex Optimization: Applications to Generalization, Memorization, and Tracing (ICML 2024 · similarity 0.86)
- Bayesian Regret Minimization in Offline Bandits (ICML 2024 · similarity 0.86)
- The Choice of Noninformative Priors for Thompson Sampling in Multiparameter Bandit Models (AAAI 2024 · similarity 0.86)
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.
