Research map / Learning theory and algorithms / Prediction and decision losses
On A Mallows-type Model For (Ranked) Choices
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 |
|---|---|---|---|
| Use RMJ instead of Mallows Smoothing or MNL for top-1 choice prediction | Top-1 sushi choice models · Sushi preference data; test display sets | log-likelihood | Helped |
| Train with a mixture of preference clusters rather than a single cluster | RMJ choice model · Top-1 sushi choice prediction | log-likelihood | Helped |
| Estimate from top-1, top-2, or top-3 feedback | RMJ model · First 10-sushi dataset | central-ranking consistency | No clear effect |
| Use the LP-Rand-Pivot speed-up heuristic | Single-cluster RMJ model · 100-sushi data | optimality gap: 1.47% average; 1.79% max | Helped |
| Use Mallows Smoothing's choice-to-ranking step before fitting a Mallows model | Mallows Smoothing heuristic · Choice data with all display sets of sizes larger than two observed infinitely often | parameter consistency | Hurt |
Closest papers by meaning
- Generalized Top-k Mallows Model for Ranked Choices (NeurIPS 2025 · similarity 0.89)
- Preference Learning Algorithms Do Not Learn Preference Rankings (NeurIPS 2024 · similarity 0.87)
- ResponseRank: Data-Efficient Reward Modeling through Preference Strength Learning (NeurIPS 2025 · similarity 0.87)
- Enhancing Preference-based Linear Bandits via Human Response Time (NeurIPS 2024 · similarity 0.86)
- Active preference learning for ordering items in- and out-of-sample (NeurIPS 2024 · similarity 0.86)
- Preference learning made easy: Everything should be understood through win rate (ICML 2025 · 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.