E. Ruben van Beesten, Nick W. Koning, David P. Morton
{"title":"评估风险规避随机程序的解决方案质量","authors":"E. Ruben van Beesten, Nick W. Koning, David P. Morton","doi":"arxiv-2408.15690","DOIUrl":null,"url":null,"abstract":"In an optimization problem, the quality of a candidate solution can be\ncharacterized by the optimality gap. For most stochastic optimization problems,\nthis gap must be statistically estimated. We show that standard estimators are\noptimistically biased for risk-averse problems, which compromises the\nstatistical guarantee on the optimality gap. We introduce estimators for\nrisk-averse problems that do not suffer from this bias. Our method relies on\nusing two independent samples, each estimating a different component of the\noptimality gap. Our approach extends a broad class of methods for estimating\nthe optimality gap from the risk-neutral case to the risk-averse case, such as\nthe multiple replications procedure and its one- and two-sample variants. Our\napproach can further make use of existing bias and variance reduction\ntechniques.","PeriodicalId":501128,"journal":{"name":"arXiv - QuantFin - Risk Management","volume":"28 1","pages":""},"PeriodicalIF":0.0000,"publicationDate":"2024-08-28","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Assessing solution quality in risk-averse stochastic programs\",\"authors\":\"E. Ruben van Beesten, Nick W. Koning, David P. Morton\",\"doi\":\"arxiv-2408.15690\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"In an optimization problem, the quality of a candidate solution can be\\ncharacterized by the optimality gap. For most stochastic optimization problems,\\nthis gap must be statistically estimated. We show that standard estimators are\\noptimistically biased for risk-averse problems, which compromises the\\nstatistical guarantee on the optimality gap. We introduce estimators for\\nrisk-averse problems that do not suffer from this bias. Our method relies on\\nusing two independent samples, each estimating a different component of the\\noptimality gap. Our approach extends a broad class of methods for estimating\\nthe optimality gap from the risk-neutral case to the risk-averse case, such as\\nthe multiple replications procedure and its one- and two-sample variants. Our\\napproach can further make use of existing bias and variance reduction\\ntechniques.\",\"PeriodicalId\":501128,\"journal\":{\"name\":\"arXiv - QuantFin - Risk Management\",\"volume\":\"28 1\",\"pages\":\"\"},\"PeriodicalIF\":0.0000,\"publicationDate\":\"2024-08-28\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"arXiv - QuantFin - Risk Management\",\"FirstCategoryId\":\"1085\",\"ListUrlMain\":\"https://doi.org/arxiv-2408.15690\",\"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 - QuantFin - Risk Management","FirstCategoryId":"1085","ListUrlMain":"https://doi.org/arxiv-2408.15690","RegionNum":0,"RegionCategory":null,"ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"","JCRName":"","Score":null,"Total":0}
Assessing solution quality in risk-averse stochastic programs
In an optimization problem, the quality of a candidate solution can be
characterized by the optimality gap. For most stochastic optimization problems,
this gap must be statistically estimated. We show that standard estimators are
optimistically biased for risk-averse problems, which compromises the
statistical guarantee on the optimality gap. We introduce estimators for
risk-averse problems that do not suffer from this bias. Our method relies on
using two independent samples, each estimating a different component of the
optimality gap. Our approach extends a broad class of methods for estimating
the optimality gap from the risk-neutral case to the risk-averse case, such as
the multiple replications procedure and its one- and two-sample variants. Our
approach can further make use of existing bias and variance reduction
techniques.