Recognizing target intent is crucial for making decisions on the battlefield.However,the imperfect and ambiguous character of battlefield situations challenges the validity and causation analysis of classical intent r...Recognizing target intent is crucial for making decisions on the battlefield.However,the imperfect and ambiguous character of battlefield situations challenges the validity and causation analysis of classical intent recognition techniques.Facing with the challenge,a target intention causal analysis paradigm is proposed by combining with an Intervention Retrieval(IR)model and a Hybrid Intention Recognition(HIR)model.The target data acquired by the sensors are modelled as Basic Probability Assignments(BPAs)based on evidence theory to create uncertain datasets.Then,the HIR model is utilized to recognize intent for a tested sample from uncertain datasets.Finally,the intervention operator under the evidence structure is utilized to perform attribute intervention on the tested sample.Data retrieval is performed in the sample database based on the IR model to generate the intention distribution of the pseudo-intervention samples to analyze the causal effects of individual sample attributes.The simulation results demonstrate that our framework successfully identifies the target intention under the evidence structure and goes further to analyze the causal impact of sample attributes on the target intention.展开更多
This paper presents the Bayes estimation and empirical Bayes estimation of causal effects in a counterfactual model. It also gives three kinds of prior distribution of the assumptions of replaceability. The experiment...This paper presents the Bayes estimation and empirical Bayes estimation of causal effects in a counterfactual model. It also gives three kinds of prior distribution of the assumptions of replaceability. The experiment shows that empirical Bayes estimation is better than other estimations when not knowing which assumption is true.展开更多
BACKGROUND Anxiety is common in patients with inflammatory bowel disease(IBD),including those with ulcerative colitis(UC)and Crohn’s disease(CD);however,the causal relationship between IBD and anxiety remains unknown...BACKGROUND Anxiety is common in patients with inflammatory bowel disease(IBD),including those with ulcerative colitis(UC)and Crohn’s disease(CD);however,the causal relationship between IBD and anxiety remains unknown.AIM To investigate the causal relationship between IBD and anxiety by using bidirectional Mendelian randomization analysis.METHODS Single nucleotide polymorphisms retrieved from genome-wide association studies(GWAS)of the European population were identified as genetic instrument variants.GWAS statistics for individuals with UC(6968 patients and 20464 controls;adults)and CD(5956 patients and 14927 controls;adults)were obtained from the International IBD Genetics Consortium.GWAS statistics for individuals with anxiety were obtained from the Psychiatric Genomics Consortium(2565 patients and 14745 controls;adults)and FinnGen project(20992 patients and 197800 controls;adults),respectively.Inverse-variance weighted was applied to assess the causal relationship,and the results were strengthened by heterogeneity,pleiotropy and leave-one-out analyses.RESULTS Genetic susceptibility to UC was associated with an increased risk of anxiety[odds ratio:1.071(95%confidence interval:1.009-1.135),P=0.023],while genetic susceptibility to CD was not associated with anxiety.Genetic susceptibility to anxiety was not associated with UC or CD.No heterogeneity or pleiotropy was observed,and the leave-one-out analysis excluded the potential influence of a particular variant.CONCLUSION This study revealed that genetic susceptibility to UC was significantly associated with anxiety and highlighted the importance of early screening for anxiety in patients with UC.展开更多
Matching is a routinely used technique to balance covariates and thereby alleviate confounding bias in causal inference with observational data.Most of the matching literatures involve the estimating of propensity sco...Matching is a routinely used technique to balance covariates and thereby alleviate confounding bias in causal inference with observational data.Most of the matching literatures involve the estimating of propensity score with parametric model,which heavily depends on the model specification.In this paper,we employ machine learning and matching techniques to learn the average causal effect.By comparing a variety of machine learning methods in terms of propensity score under extensive scenarios,we find that the ensemble methods,especially generalized random forests,perform favorably with others.We apply all the methods to the data of tropical storms that occurred on the mainland of China since 1949.展开更多
Counterfactual model is put forward to discuss the causal inference in the directed acyclic graph and its corresponding identifiability is thus studied with the ancillary information based on conditional independence....Counterfactual model is put forward to discuss the causal inference in the directed acyclic graph and its corresponding identifiability is thus studied with the ancillary information based on conditional independence. It is shown that the assumption of ignorability can be expanded to the assumption of replaceability, under which the causal effects are identifiable.展开更多
We consider the estimation of causal treatment effect using nonparametric regression orinverse propensity weighting together with sufficient dimension reduction for searching lowdimensional covariate subsets. A specia...We consider the estimation of causal treatment effect using nonparametric regression orinverse propensity weighting together with sufficient dimension reduction for searching lowdimensional covariate subsets. A special case of this problem is the estimation of a responseeffect with data having ignorable missing response values. An issue that is not well addressedin the literature is whether the estimation of the low-dimensional covariate subsets by sufficient dimension reduction has an impact on the asymptotic variance of the resulting causaleffect estimator. With some incorrect or inaccurate statements, many researchers believe thatthe estimation of the low-dimensional covariate subsets by sufficient dimension reduction doesnot affect the asymptotic variance. We rigorously establish a result showing that this is nottrue unless the low-dimensional covariate subsets include some covariates superfluous for estimation, and including such covariates loses efficiency. Our theory is supplemented by somesimulation results.展开更多
BACKGROUND Numerous observational studies have documented a correlation between inflammatory bowel disease(IBD)and an increased risk of dementia.However,the causality of their associations remains elusive.AIM To asses...BACKGROUND Numerous observational studies have documented a correlation between inflammatory bowel disease(IBD)and an increased risk of dementia.However,the causality of their associations remains elusive.AIM To assess the causal relationship between IBD and the occurrence of all-cause dementia using the two-sample Mendelian randomization(MR)method.METHODS Genetic variants extracted from the large genome-wide association study(GWAS)for IBD(the International IBD Genetics Consortium,n=34652)were used to identify the causal link between IBD and dementia(FinnGen,n=306102).The results of the study were validated via another IBD GWAS(United Kingdom Biobank,n=463372).Moreover,MR egger intercept,MR pleiotropy residual sum and outlier,and Cochran's Q test were employed to evaluate pleiotropy and heterogeneity.Finally,multiple MR methods were performed to estimate the effects of genetically predicted IBD on dementia,with the inverse variance weighted approach adopted as the primary analysis.RESULTS The results of the pleiotropy and heterogeneity tests revealed an absence of significant pleiotropic effects or heterogeneity across all genetic variants in outcome GWAS.No evidence of a causal effect between IBD and the risk of dementia was identified in the inverse variance weighted[odds ratio(OR)=0.980,95%CI:0.942-1.020,P value=0.325],weighted median(OR=0.964,95%CI:0.914-1.017,P value=0.180),and MR-Egger(OR=0.963,95%CI:0.867-1.070,P value=0.492)approaches.Consistent results were observed in validation analyses.Reverse MR analysis also showed no effect of dementia on the development of IBD.Furthermore,MR analysis suggested that IBD and its subtypes did not causally affect allcause dementia and its four subtypes,including dementia in Alzheimer's disease,vascular dementia,dementia in other diseases classified elsewhere,and unspecified dementia.CONCLUSION Taken together,our MR study signaled that IBD and its subentities were not genetically associated with all-cause dementia or its subtypes.Further large prospective studies are warranted to elucidate the impact of intestinal inflammation on the development of dementia.展开更多
反事实预测和选择偏差是因果效应估计中的重大挑战。为对潜在协变量的复杂混杂分布进行有效表征,同时增强反事实预测泛化能力,提出一种面向工业因果效应估计应用的重加权对抗变分自编码器网络(RVAENet)模型。针对混杂分布去偏问题,借鉴...反事实预测和选择偏差是因果效应估计中的重大挑战。为对潜在协变量的复杂混杂分布进行有效表征,同时增强反事实预测泛化能力,提出一种面向工业因果效应估计应用的重加权对抗变分自编码器网络(RVAENet)模型。针对混杂分布去偏问题,借鉴域适应思想,采用对抗学习机制对由变分自编码器(VAE)获得的隐含变量进行表示学习的分布平衡;在此基础上,通过学习样本倾向性权重对样本进行重加权,进一步缩小实验组(Treatment)与对照组(Control)样本间的分布差异。实验结果表明,在工业真实场景数据集的两个场景下,所提模型的提升曲线下的面积(AUUC)比TEDVAE(Treatment Effect with Disentangled VAE)分别提升了15.02%、16.02%;在公开数据集上,所提模型的平均干预效果(ATE)和异构估计精度(PEHE)普遍取得最优结果。展开更多
目的采用两样本孟德尔随机化(MR)探讨417种循环代谢产物与吉兰-巴雷综合征(GBS)风险的因果关系。方法通过MRC IEU OpenGWAS项目获得3个循环代谢产物全基因组关联研究(GWAS)数据,分别为“met-a”“met-c”和“met-d”。GBS相关的单核苷...目的采用两样本孟德尔随机化(MR)探讨417种循环代谢产物与吉兰-巴雷综合征(GBS)风险的因果关系。方法通过MRC IEU OpenGWAS项目获得3个循环代谢产物全基因组关联研究(GWAS)数据,分别为“met-a”“met-c”和“met-d”。GBS相关的单核苷酸多态性位点(SNPs)数据来源于芬兰生物银行数据库,表型代码为“finn-b-G6_GUILBAR”。将来自GWAS的与循环代谢产物密切相关的遗传变异数据(SNPs)作为工具变量(IVs),与来自芬兰的GBS GWAS数据进行双样本MR分析,主要采用随机效应模型的逆方差加权(IVW)方法,根据效应指标优势比(OR)和95%CI评估结果。使用留一法、异质性检验、水平基因多效性检验验证结果的稳定性和可靠性。结果共5种循环代谢产物具有与GBS因果关系的提示性证据。其中,肌酐(OR=2.924,95%CI:1.194~7.163,P=0.019)、谷氨酰胺(OR=1.902,95%CI:1.007~3.592,P=0.048)、异戊酰基肉碱(OR=140.767,95%CI:3.510~5645.336,P=0.009)的循环水平与GBS风险较高有关。相反,基因预测葡萄糖(OR=0.308,95%CI:0.010~0.981,P=0.046)、X-11491(OR=0.069,95%CI:0.007~0.707,P=0.024)的循环水平与GBS风险呈负相关。结论肌酐、谷氨酰胺、异戊酰基肉碱、葡萄糖、X-11491可能与GBS有因果关系。展开更多
基金partially supported by the National Natural Science Foundation of China(No.62173272)。
文摘Recognizing target intent is crucial for making decisions on the battlefield.However,the imperfect and ambiguous character of battlefield situations challenges the validity and causation analysis of classical intent recognition techniques.Facing with the challenge,a target intention causal analysis paradigm is proposed by combining with an Intervention Retrieval(IR)model and a Hybrid Intention Recognition(HIR)model.The target data acquired by the sensors are modelled as Basic Probability Assignments(BPAs)based on evidence theory to create uncertain datasets.Then,the HIR model is utilized to recognize intent for a tested sample from uncertain datasets.Finally,the intervention operator under the evidence structure is utilized to perform attribute intervention on the tested sample.Data retrieval is performed in the sample database based on the IR model to generate the intention distribution of the pseudo-intervention samples to analyze the causal effects of individual sample attributes.The simulation results demonstrate that our framework successfully identifies the target intention under the evidence structure and goes further to analyze the causal impact of sample attributes on the target intention.
文摘This paper presents the Bayes estimation and empirical Bayes estimation of causal effects in a counterfactual model. It also gives three kinds of prior distribution of the assumptions of replaceability. The experiment shows that empirical Bayes estimation is better than other estimations when not knowing which assumption is true.
基金Supported by China Postdoctoral Science Foundation,No.2021M701614Guangdong Basic and Applied Basic Research Foundation,No.2022A1515111063,No.2022A1515111045Foundation of Guangdong Provincial People’s Hospital,No.8200010545。
文摘BACKGROUND Anxiety is common in patients with inflammatory bowel disease(IBD),including those with ulcerative colitis(UC)and Crohn’s disease(CD);however,the causal relationship between IBD and anxiety remains unknown.AIM To investigate the causal relationship between IBD and anxiety by using bidirectional Mendelian randomization analysis.METHODS Single nucleotide polymorphisms retrieved from genome-wide association studies(GWAS)of the European population were identified as genetic instrument variants.GWAS statistics for individuals with UC(6968 patients and 20464 controls;adults)and CD(5956 patients and 14927 controls;adults)were obtained from the International IBD Genetics Consortium.GWAS statistics for individuals with anxiety were obtained from the Psychiatric Genomics Consortium(2565 patients and 14745 controls;adults)and FinnGen project(20992 patients and 197800 controls;adults),respectively.Inverse-variance weighted was applied to assess the causal relationship,and the results were strengthened by heterogeneity,pleiotropy and leave-one-out analyses.RESULTS Genetic susceptibility to UC was associated with an increased risk of anxiety[odds ratio:1.071(95%confidence interval:1.009-1.135),P=0.023],while genetic susceptibility to CD was not associated with anxiety.Genetic susceptibility to anxiety was not associated with UC or CD.No heterogeneity or pleiotropy was observed,and the leave-one-out analysis excluded the potential influence of a particular variant.CONCLUSION This study revealed that genetic susceptibility to UC was significantly associated with anxiety and highlighted the importance of early screening for anxiety in patients with UC.
基金supported by the National Key Research and Development Program of China Grant 2017YFA0604903National Natural Science Foundation of China Grant(Nos.11671338,11971064)。
文摘Matching is a routinely used technique to balance covariates and thereby alleviate confounding bias in causal inference with observational data.Most of the matching literatures involve the estimating of propensity score with parametric model,which heavily depends on the model specification.In this paper,we employ machine learning and matching techniques to learn the average causal effect.By comparing a variety of machine learning methods in terms of propensity score under extensive scenarios,we find that the ensemble methods,especially generalized random forests,perform favorably with others.We apply all the methods to the data of tropical storms that occurred on the mainland of China since 1949.
基金This work was supported by the National Natural Science Foundation of China (Grant Nos. 39930160, 19871003).
文摘Counterfactual model is put forward to discuss the causal inference in the directed acyclic graph and its corresponding identifiability is thus studied with the ancillary information based on conditional independence. It is shown that the assumption of ignorability can be expanded to the assumption of replaceability, under which the causal effects are identifiable.
基金This research was partially supported through a PatientCentered Outcomes Research Institute(PCORI)Award(ME-1409-21219)This research was also supported by the National Natural Science Foundation of China(11501208)+2 种基金Fundamental Research Funds for the Central Universities,National Social Science Foundation(13BTJ009)the Chinese 111 Project grant(B14019)the U.S.National Science Foundation(DMS-1305474 and DMS-1612873).
文摘We consider the estimation of causal treatment effect using nonparametric regression orinverse propensity weighting together with sufficient dimension reduction for searching lowdimensional covariate subsets. A special case of this problem is the estimation of a responseeffect with data having ignorable missing response values. An issue that is not well addressedin the literature is whether the estimation of the low-dimensional covariate subsets by sufficient dimension reduction has an impact on the asymptotic variance of the resulting causaleffect estimator. With some incorrect or inaccurate statements, many researchers believe thatthe estimation of the low-dimensional covariate subsets by sufficient dimension reduction doesnot affect the asymptotic variance. We rigorously establish a result showing that this is nottrue unless the low-dimensional covariate subsets include some covariates superfluous for estimation, and including such covariates loses efficiency. Our theory is supplemented by somesimulation results.
文摘BACKGROUND Numerous observational studies have documented a correlation between inflammatory bowel disease(IBD)and an increased risk of dementia.However,the causality of their associations remains elusive.AIM To assess the causal relationship between IBD and the occurrence of all-cause dementia using the two-sample Mendelian randomization(MR)method.METHODS Genetic variants extracted from the large genome-wide association study(GWAS)for IBD(the International IBD Genetics Consortium,n=34652)were used to identify the causal link between IBD and dementia(FinnGen,n=306102).The results of the study were validated via another IBD GWAS(United Kingdom Biobank,n=463372).Moreover,MR egger intercept,MR pleiotropy residual sum and outlier,and Cochran's Q test were employed to evaluate pleiotropy and heterogeneity.Finally,multiple MR methods were performed to estimate the effects of genetically predicted IBD on dementia,with the inverse variance weighted approach adopted as the primary analysis.RESULTS The results of the pleiotropy and heterogeneity tests revealed an absence of significant pleiotropic effects or heterogeneity across all genetic variants in outcome GWAS.No evidence of a causal effect between IBD and the risk of dementia was identified in the inverse variance weighted[odds ratio(OR)=0.980,95%CI:0.942-1.020,P value=0.325],weighted median(OR=0.964,95%CI:0.914-1.017,P value=0.180),and MR-Egger(OR=0.963,95%CI:0.867-1.070,P value=0.492)approaches.Consistent results were observed in validation analyses.Reverse MR analysis also showed no effect of dementia on the development of IBD.Furthermore,MR analysis suggested that IBD and its subtypes did not causally affect allcause dementia and its four subtypes,including dementia in Alzheimer's disease,vascular dementia,dementia in other diseases classified elsewhere,and unspecified dementia.CONCLUSION Taken together,our MR study signaled that IBD and its subentities were not genetically associated with all-cause dementia or its subtypes.Further large prospective studies are warranted to elucidate the impact of intestinal inflammation on the development of dementia.
文摘反事实预测和选择偏差是因果效应估计中的重大挑战。为对潜在协变量的复杂混杂分布进行有效表征,同时增强反事实预测泛化能力,提出一种面向工业因果效应估计应用的重加权对抗变分自编码器网络(RVAENet)模型。针对混杂分布去偏问题,借鉴域适应思想,采用对抗学习机制对由变分自编码器(VAE)获得的隐含变量进行表示学习的分布平衡;在此基础上,通过学习样本倾向性权重对样本进行重加权,进一步缩小实验组(Treatment)与对照组(Control)样本间的分布差异。实验结果表明,在工业真实场景数据集的两个场景下,所提模型的提升曲线下的面积(AUUC)比TEDVAE(Treatment Effect with Disentangled VAE)分别提升了15.02%、16.02%;在公开数据集上,所提模型的平均干预效果(ATE)和异构估计精度(PEHE)普遍取得最优结果。
文摘目的采用两样本孟德尔随机化(MR)探讨417种循环代谢产物与吉兰-巴雷综合征(GBS)风险的因果关系。方法通过MRC IEU OpenGWAS项目获得3个循环代谢产物全基因组关联研究(GWAS)数据,分别为“met-a”“met-c”和“met-d”。GBS相关的单核苷酸多态性位点(SNPs)数据来源于芬兰生物银行数据库,表型代码为“finn-b-G6_GUILBAR”。将来自GWAS的与循环代谢产物密切相关的遗传变异数据(SNPs)作为工具变量(IVs),与来自芬兰的GBS GWAS数据进行双样本MR分析,主要采用随机效应模型的逆方差加权(IVW)方法,根据效应指标优势比(OR)和95%CI评估结果。使用留一法、异质性检验、水平基因多效性检验验证结果的稳定性和可靠性。结果共5种循环代谢产物具有与GBS因果关系的提示性证据。其中,肌酐(OR=2.924,95%CI:1.194~7.163,P=0.019)、谷氨酰胺(OR=1.902,95%CI:1.007~3.592,P=0.048)、异戊酰基肉碱(OR=140.767,95%CI:3.510~5645.336,P=0.009)的循环水平与GBS风险较高有关。相反,基因预测葡萄糖(OR=0.308,95%CI:0.010~0.981,P=0.046)、X-11491(OR=0.069,95%CI:0.007~0.707,P=0.024)的循环水平与GBS风险呈负相关。结论肌酐、谷氨酰胺、异戊酰基肉碱、葡萄糖、X-11491可能与GBS有因果关系。
