Research map / Learning theory and algorithms / Prediction and decision losses
Regression under Human Assistance
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 |
|---|---|---|---|
| greedy algorithm with alpha-submodular guarantees vs baselines (DS, distorted greedy, Triage, CRR) | ridge regression under human assistance · Synthetic data | held-out MSE | Helped |
| greedy algorithm vs baselines (DS, distorted greedy, Triage, CRR) | ridge regression under human assistance · Real data (Messidor, Stare-H, Stare-D, Hatespeech) | held-out MSE | Helped |
| outsource a fraction of samples to humans instead of full automation | ridge regression · Synthetic data | held-out MSE | Helped |
| outsource samples when human error is high (ablation) | greedy algorithm · Synthetic data | held-out MSE | Hurt |
| outsource more samples than the fraction humans predict well (ablation) | greedy algorithm · Messidor, Stare-H, Stare-D | held-out MSE | Hurt |
| use nearest neighbors or MLP instead of logistic regression as additional model h_theta (ablation) | greedy algorithm · Synthetic logistic dataset | held-out MSE | Helped |
| use logistic regression instead of NN as additional model h_theta (ablation) | greedy algorithm · Real data | held-out MSE | Helped |
| vary ground set size, embedding dimension and PCA dimension (ablation) | greedy algorithm · Hatespeech and Messidor | held-out MSE | No clear effect |
Closest papers by meaning
- Classification Under Human Assistance (AAAI 2021 · similarity 0.89)
- A Sharper Generalization Bound for Divide-and-Conquer Ridge Regression (AAAI 2019 · similarity 0.86)
- ElliCE: Efficient and Provably Robust Algorithmic Recourse via the Rashomon Sets (NeurIPS 2025 · similarity 0.85)
- Piecewise Constant and Linear Regression Trees: An Optimal Dynamic Programming Approach (ICML 2024 · similarity 0.85)
- Learning to Decide with AI Assistance under Human-Alignment (NeurIPS 2026 · similarity 0.84)
- Corporate Needs You to Find the Difference: Revisiting Submodular and Supermodular Ratio Optimization Problems (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.