Lucas Kock, G. S. Rodrigues, Scott A. Sisson, Nadja Klein, David J. Nott
{"title":"使用本地化 PIT 映射的校准多元回归","authors":"Lucas Kock, G. S. Rodrigues, Scott A. Sisson, Nadja Klein, David J. Nott","doi":"arxiv-2409.10855","DOIUrl":null,"url":null,"abstract":"Calibration ensures that predicted uncertainties align with observed\nuncertainties. While there is an extensive literature on recalibration methods\nfor univariate probabilistic forecasts, work on calibration for multivariate\nforecasts is much more limited. This paper introduces a novel post-hoc\nrecalibration approach that addresses multivariate calibration for potentially\nmisspecified models. Our method involves constructing local mappings between\nvectors of marginal probability integral transform values and the space of\nobservations, providing a flexible and model free solution applicable to\ncontinuous, discrete, and mixed responses. We present two versions of our\napproach: one uses K-nearest neighbors, and the other uses normalizing flows.\nEach method has its own strengths in different situations. We demonstrate the\neffectiveness of our approach on two real data applications: recalibrating a\ndeep neural network's currency exchange rate forecast and improving a\nregression model for childhood malnutrition in India for which the multivariate\nresponse has both discrete and continuous components.","PeriodicalId":501425,"journal":{"name":"arXiv - STAT - Methodology","volume":"44 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-09-17","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Calibrated Multivariate Regression with Localized PIT Mappings\",\"authors\":\"Lucas Kock, G. S. Rodrigues, Scott A. Sisson, Nadja Klein, David J. Nott\",\"doi\":\"arxiv-2409.10855\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"Calibration ensures that predicted uncertainties align with observed\\nuncertainties. While there is an extensive literature on recalibration methods\\nfor univariate probabilistic forecasts, work on calibration for multivariate\\nforecasts is much more limited. This paper introduces a novel post-hoc\\nrecalibration approach that addresses multivariate calibration for potentially\\nmisspecified models. Our method involves constructing local mappings between\\nvectors of marginal probability integral transform values and the space of\\nobservations, providing a flexible and model free solution applicable to\\ncontinuous, discrete, and mixed responses. We present two versions of our\\napproach: one uses K-nearest neighbors, and the other uses normalizing flows.\\nEach method has its own strengths in different situations. We demonstrate the\\neffectiveness of our approach on two real data applications: recalibrating a\\ndeep neural network's currency exchange rate forecast and improving a\\nregression model for childhood malnutrition in India for which the multivariate\\nresponse has both discrete and continuous components.\",\"PeriodicalId\":501425,\"journal\":{\"name\":\"arXiv - STAT - Methodology\",\"volume\":\"44 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-09-17\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - STAT - Methodology\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2409.10855\",\"RegionNum\":0,\"RegionCategory\":null,\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"\",\"JCRName\":\"\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"arXiv - STAT - Methodology","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2409.10855","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
引用次数: 0
摘要
校准可确保预测的不确定性与观测到的不确定性相一致。关于单变量概率预测的重新校准方法已有大量文献,但关于多变量预测的校准工作则有限得多。本文介绍了一种新颖的事后重新校准方法,可解决潜在不确定模型的多变量校准问题。我们的方法涉及在边际概率积分变换值向量和观测空间之间构建局部映射,提供一种灵活的、不受模型限制的解决方案,适用于连续、离散和混合响应。我们介绍了我们方法的两个版本:一个使用 K 最近邻,另一个使用归一化流。我们在两个实际数据应用中展示了我们方法的有效性:重新校准深度神经网络的汇率预测,以及改进印度儿童营养不良的回归模型,其中多变量响应既有离散成分,也有连续成分。
Calibrated Multivariate Regression with Localized PIT Mappings
Calibration ensures that predicted uncertainties align with observed
uncertainties. While there is an extensive literature on recalibration methods
for univariate probabilistic forecasts, work on calibration for multivariate
forecasts is much more limited. This paper introduces a novel post-hoc
recalibration approach that addresses multivariate calibration for potentially
misspecified models. Our method involves constructing local mappings between
vectors of marginal probability integral transform values and the space of
observations, providing a flexible and model free solution applicable to
continuous, discrete, and mixed responses. We present two versions of our
approach: one uses K-nearest neighbors, and the other uses normalizing flows.
Each method has its own strengths in different situations. We demonstrate the
effectiveness of our approach on two real data applications: recalibrating a
deep neural network's currency exchange rate forecast and improving a
regression model for childhood malnutrition in India for which the multivariate
response has both discrete and continuous components.