Research map / Learning theory and algorithms / Sample complexity
The Limits of Learning with Missing Data
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 |
|---|---|---|---|
| Restrict learner to observe at most one attribute per example (vs. two) for squared-loss Lasso/Ridge regression | Linear regression with squared loss (Lasso/Ridge) · Limited attribute observation model, theoretical lower bound | achievable precision (epsilon-optimality) | Hurt |
| Restrict learner to observe at most d-1 attributes per example (vs. all d) for absolute-loss regression | Linear regression with absolute loss · Limited attribute observation model, theoretical lower bound with d >= 4 | achievable precision (epsilon-optimality) | Hurt |
| Restrict learner to observe at most d-1 attributes per example for hinge-loss classification (SVM) | Soft-margin SVM (hinge loss) · Limited attribute observation model, theoretical lower bound | achievable precision (epsilon-optimality) | Hurt |
| Use general algorithm (Algorithm 1) observing k attributes per example, with m = d/(d-k) samples | Regression/classification with missing attributes · Theoretical upper bound | expected excess loss | No clear effect |
Closest papers by meaning
- Regression Learning with Limited Observations of Multivariate Outcomes and Features (ICML 2024 · similarity 0.87)
- Random features models: a way to study the success of naive imputation (ICML 2024 · similarity 0.86)
- Prediction models that learn to avoid missing values (ICML 2025 · similarity 0.85)
- How Much is Unseen Depends Chiefly on Information About the Seen (ICLR 2025 · similarity 0.84)
- Distribution Shift in Missing Data Imputation: A Risk-Based Perspective and Importance-Weighted Correction under MAR (NeurIPS 2026 · similarity 0.83)
- Learning from Interval Targets (NeurIPS 2025 · 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.