Research map / Learning theory and algorithms / Matrix and tensor methods
Consistent Plug-in Classifiers for Complex Objectives and Constraints
Shiv Kumar Tavker, Harish Guruprasad Ramaswamy, Harikrishna Narasimhan
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 proposed method instead of unconstrained optimization | Q-mean objective · Lawschool, DP<=0.05 | constraint violation: 0.25 -> 0.02 | Improved |
| use proposed method instead of Error-Con | Q-mean objective · Adult, DP<=0.05 | Q-mean: 0.30 -> 0.18 | Improved |
| use proposed method instead of COCO | Q-mean objective · Adult, DP<=0.05 | Q-mean: 0.31 -> 0.18 | Improved |
| use proposed method instead of 3-Player | Q-mean objective · Crimes, DP<=0.05 | Q-mean: 0.24 -> 0.22 | Improved |
| use proposed method instead of 3-Player | Q-mean objective · Lawschool, DP<=0.05 | Q-mean: 0.24 -> 0.25 | Hurt |
| use proposed method instead of 3-Player | G-mean objective · Lawschool, EO<=0.05 | G-mean: 0.22 -> 0.26 | Hurt |
| use proposed method instead of COCO | G-mean objective · Crimes, EO<=0.05 | equal opportunity violation: 0.06 -> 0.03 | Improved |
| use proposed method instead of 3-Player | H-mean objective · Default, coverage>=0.25 | H-mean: 0.42 -> 0.34 | Improved |
| use proposed method instead of 3-Player | H-mean objective · Lawschool, coverage>=0.25 | H-mean: 0.56 -> 0.55 | No clear effect |
| use proposed method instead of COCO | G-mean with equal opportunity · COMPAS convergence | plug-in calls to zero violation: fewer calls (100) | Improved |
| use proposed method instead of COCO | G-mean with equal opportunity · Crimes and Law School convergence | training violation: COCO fails after 2000 calls | Improved |
| use proposed method instead of COCO | G-mean with equal opportunity · Adult convergence | convergence speed | Hurt |
Closest papers by meaning
- Consistent Multiclass Algorithms for Complex Metrics and Constraints (JMLR 2024 · similarity 0.89)
- Learning Sparse Classifiers: Continuous and Mixed Integer Optimization Perspectives (JMLR 2021 · similarity 0.83)
- On Rate-Optimal Partitioning Classification from Observable and from Privatised Data (NeurIPS 2026 · similarity 0.83)
- Consistent algorithms for multi-label classification with macro-at-$k$ metrics (ICLR 2024 · similarity 0.83)
- Learning with Statistical Equality Constraints (NeurIPS 2025 · similarity 0.82)
- Trading off Consistency and Dimensionality of Convex Surrogates for Multiclass Classification (NeurIPS 2024 · similarity 0.82)
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.
