AI Paper Atlas

Research map / Learning theory and algorithms / Prediction and decision losses

Limits and Potential of Score-Based Data Valuation: Redundancy, Complementarity, and Non-Monotonicity

Kumar Kshitij Patel, Sai Praneeth Karimireddy, Raul Fernandez, Manolis Zampetakis

NeurIPS 2026 · Learning theory and algorithms › Prediction and decision losses · approach: Shapley value attribution

Read the paperCheck your idea against 137,217 papers

AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Analyzes when score-based data valuation methods (Shapley, LOO) reliably guide downstream data decisions, identifying redundancy, complementarity, and non-monotonicity as governing structural properties.

Problem

Score-based data valuation methods such as Shapley values and leave-one-out are used for attribution, selection, and pricing, but it is unclear when their scalar scores reliably guide downstream decisions.

Approach

Argues that the success of score-based valuation is governed by three structural properties of the learning problem (substitutability, complementarity, and non-monotonicity), characterizing when scalar scores are reliable; full text unavailable so details beyond the abstract are not established.

Experimental setup

Builds on: Shapley values leave-one-out

Closest papers by meaning

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.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy