油中溶解气体分析(dissolved gas analysis,DGA)技术可以有效发现充油变压器内部的早期潜伏性故障,是对充油设备进行诊断的常用检测方法。CART(classification and regression trees)分类回归树是一种可处理连续属性的不平衡算法。SAMM...油中溶解气体分析(dissolved gas analysis,DGA)技术可以有效发现充油变压器内部的早期潜伏性故障,是对充油设备进行诊断的常用检测方法。CART(classification and regression trees)分类回归树是一种可处理连续属性的不平衡算法。SAMME作为Ada Boost二分类算法的延伸,可将多个略好于随机猜测的弱分类器组合提升为分类精度更高的强分类器,完成多分类任务。针对单一算法往往不能满足实际工程对分类精度的需求且高精度算法难以获得的问题,采用SAMME算法对CART弱学习器的权重根据误差不断调整,再通过加权投票将其组合提升为强分类器,并采用V折交叉验证确定SAMME-CART模型的最优迭代次数。以此提高故障诊断模型的泛化能力。实验结果表明,运用该模型结合DGA技术对变压器故障进行诊断,相比于单一CART算法,诊断准确率整体提高了18.7%,表明该方法是可行的。展开更多
Background and Aims:Previous studies have reported that the single nucleotide polymorphisms(SNPs)of SAMM50-rs738491,PARVB-rs5764455 and PNPLA3-rs738409 are associated with nonalcoholic fatty liver disease(NAFLD).Howev...Background and Aims:Previous studies have reported that the single nucleotide polymorphisms(SNPs)of SAMM50-rs738491,PARVB-rs5764455 and PNPLA3-rs738409 are associated with nonalcoholic fatty liver disease(NAFLD).However,no studies have examined the effect of interactions between these three genotypes to affect liver disease severity.We assessed the effect of these three SNPs on nonalcoholic steatohepatitis(NASH)and also examined the gene-gene interactions in a Chinese population with biopsy-confirmed NAFLD.Methods:We enrolled 415 consecutive adult individuals with biopsy-proven NAFLD.Multivariable logistic regres-sion analysis was undertaken to test associations between NASH and SNPs in SAMM50-rs738491,PARVB-rs5764455 and PNPLA3-rs738409.Gene-gene interactions were ana-lyzed by performing a generalized multifactor dimensionality reduction(GMDR)analysis.Results:The mean±standard deviation age of these 415 patients was 41.3±12.5 years,and 75.9%were men.Patients with SAMM50-rs738491 TT,PARVB-rs5764455 AA or PNPLA3-rs738409 GG genotypes had a higher risk of NASH,even after adjustment for age,sex and body mass index.GMDR analysis showed that the combination of all three SNPs was the best model for predicting NASH.Additionally,the odds ratio of the haplotype T-A-G for predicting the risk of NASH was nearly three times higher than that of the haplotype G-C-C.Conclusions:NAFLD patients carrying the SAMM50-rs738491 TT,PARVB-rs5764455 AA or PNPLA3-rs738409 GG genotypes are at greater risk of NASH.These three SNPs may synergistically interact to increase susceptibility to NASH.展开更多
文摘油中溶解气体分析(dissolved gas analysis,DGA)技术可以有效发现充油变压器内部的早期潜伏性故障,是对充油设备进行诊断的常用检测方法。CART(classification and regression trees)分类回归树是一种可处理连续属性的不平衡算法。SAMME作为Ada Boost二分类算法的延伸,可将多个略好于随机猜测的弱分类器组合提升为分类精度更高的强分类器,完成多分类任务。针对单一算法往往不能满足实际工程对分类精度的需求且高精度算法难以获得的问题,采用SAMME算法对CART弱学习器的权重根据误差不断调整,再通过加权投票将其组合提升为强分类器,并采用V折交叉验证确定SAMME-CART模型的最优迭代次数。以此提高故障诊断模型的泛化能力。实验结果表明,运用该模型结合DGA技术对变压器故障进行诊断,相比于单一CART算法,诊断准确率整体提高了18.7%,表明该方法是可行的。
基金supported by grants from the National Natural Science Foundation of China(82070588)the High Level Creative Talents from Department of Public Health in Zhejiang Proince(S2032102600032)+2 种基金Project of New Century 551 Talent Nurturing in WenzhouGT was supported in part by grants from the School of Medicine,University of Verona,Verona,ItalyCDB was supported in part by the Southampton NIHR Biomedical Research Centre(IS-BRC-20004),UK.
文摘Background and Aims:Previous studies have reported that the single nucleotide polymorphisms(SNPs)of SAMM50-rs738491,PARVB-rs5764455 and PNPLA3-rs738409 are associated with nonalcoholic fatty liver disease(NAFLD).However,no studies have examined the effect of interactions between these three genotypes to affect liver disease severity.We assessed the effect of these three SNPs on nonalcoholic steatohepatitis(NASH)and also examined the gene-gene interactions in a Chinese population with biopsy-confirmed NAFLD.Methods:We enrolled 415 consecutive adult individuals with biopsy-proven NAFLD.Multivariable logistic regres-sion analysis was undertaken to test associations between NASH and SNPs in SAMM50-rs738491,PARVB-rs5764455 and PNPLA3-rs738409.Gene-gene interactions were ana-lyzed by performing a generalized multifactor dimensionality reduction(GMDR)analysis.Results:The mean±standard deviation age of these 415 patients was 41.3±12.5 years,and 75.9%were men.Patients with SAMM50-rs738491 TT,PARVB-rs5764455 AA or PNPLA3-rs738409 GG genotypes had a higher risk of NASH,even after adjustment for age,sex and body mass index.GMDR analysis showed that the combination of all three SNPs was the best model for predicting NASH.Additionally,the odds ratio of the haplotype T-A-G for predicting the risk of NASH was nearly three times higher than that of the haplotype G-C-C.Conclusions:NAFLD patients carrying the SAMM50-rs738491 TT,PARVB-rs5764455 AA or PNPLA3-rs738409 GG genotypes are at greater risk of NASH.These three SNPs may synergistically interact to increase susceptibility to NASH.
