提出新方法纠正缺失数据中的分布偏移问题，提升估计准确性。
Distribution Shift in Missing Data Imputation: A Risk-Based Perspective and Importance-Weighted Correction under MAR
- 将插补建模为风险最小化问题，识别出训练与评估分布差异
- 模拟实验显示均方根误差降低3%，沃尔德斯坦距离减少7%
- 适合处理依赖数据的缺失场景，对医疗、金融等高风险领域有实用价值
缺失数据插补是机器学习中的基础问题，模型基于观测数据估计未观测值。本文将插补模型学习严格建模为均方误差风险最小化问题。当缺失概率依赖于数据时，现有许多先进方法未能考虑由此产生的观测数据分布与完整数据分布之间的分布偏移，导致其无法在完整数据分布上最小化均方误差。为此，我们提出一种新插补算法，从观测数据中学习插补模型时显式地校正该分布偏移。模拟研究表明，在与原始基线相同条件下，该方法实现了稳定的性能提升，平均使均方根误差降低3%，沃尔德斯坦距离减少7%。
原文摘要 · Abstract (English)
Missing data imputation, where a model is trained on observed data to estimate unobserved values, is a fundamental problem in machine learning. In this paper, we rigorously formulate imputation model learning as a mean-squared error risk minimisation problem. We show that when the probability of missingness depends on the data, many state-of-the-art methods fail to account for the resulting distribution shift between the observed data used for training and the full data distribution used for evaluation. Consequently, these approaches do not minimise mean-squared error on the full data distribution. Instead, we propose a novel imputation algorithm designed to learn an imputation model from the observed data while explicitly accounting for this distribution shift. Simulation studies show consistent improvements over otherwise identical uncorrected baselines, with average reductions of 3% in RMSE and 7% in Wasserstein distance.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。