Although high-dose methotrexate(HD-MTX)is the most effective drug against primary CNS lymphomas(PCNSL),outcome-de-termining variables related to its administration schedule have not been defined.The impact on toxicity...Although high-dose methotrexate(HD-MTX)is the most effective drug against primary CNS lymphomas(PCNSL),outcome-de-termining variables related to its administration schedule have not been defined.The impact on toxicity and outcome of the area under thecurve(AUC(MTX)),dose intensity(DI(MTX))and infusion rate(IR(MTX))of MTX and plsamatic creatinine clearance(CL(crea))was investigated in a retrospective series of 45 PCNSL patients treated with three different HD-MTX-basedcombinations.Anticon-vulsants were administered in 31 pts(69%).Age>60 years,anticonvulsant therapy,slow IR(MTX)(</=800 mgm(-2)h(-1)),and reduced DI(MTX)(</=1000 mgm(-2)wk(-1))were significantly correlated with low AUC(MTX)values.Seven pa-展开更多
In diagnostic trials, clustered data are obtained when several subunits of the same patient are observed. Within-cluster correlations need to be taken into account when analyzing such clustered data. A nonparametric m...In diagnostic trials, clustered data are obtained when several subunits of the same patient are observed. Within-cluster correlations need to be taken into account when analyzing such clustered data. A nonparametric method has been proposed by Obuchowski (1997) to estimate the Receiver Operating Characteristic curve area (AUC) for such clustered data. However, Obuchowski’s estimator gives equal weight to all pairwise rankings within and between cluster. In this paper, we modify Obuchowski’s estimate by allowing weights for the pairwise rankings vary across clusters. We consider the optimal weights for estimating one AUC as well as two AUCs’ difference. Our results in this paper show that the optimal weights depends on not only the within-patient correlation but also the proportion of patients that have both unaffected and affected units. More importantly, we show that the loss of efficiency using equal weight instead of our optimal weights can be severe when there is a large within-cluster correlation and the proportion of patients that have both unaffected and affected units is small.展开更多
接收者操作特性(Receiver operating characteristics,ROC)曲线下面积(Area under the ROC curve,AUC)常被用于度量分类器在整个类先验分布上的总体分类性能.原始Boosting算法优化分类精度,但在AUC度量下并非最优.提出了一种AUC优化Boos...接收者操作特性(Receiver operating characteristics,ROC)曲线下面积(Area under the ROC curve,AUC)常被用于度量分类器在整个类先验分布上的总体分类性能.原始Boosting算法优化分类精度,但在AUC度量下并非最优.提出了一种AUC优化Boosting改进算法,通过在原始Boosting迭代中引入数据重平衡操作,实现弱学习算法优化目标从精度向AUC的迁移.实验结果表明,较之原始Boosting算法,新算法在AUC度量下能获得更好性能.展开更多
针对借贷过程中的信息不对称问题,为更有效地整合不同的数据源和贷款违约预测模型,提出一种集成学习的训练方法,使用AUC(Area Under Curve)值和Q统计值对学习器的准确性和多样性进行度量,并实现了基于AUC和Q统计值的集成学习训练算法(TA...针对借贷过程中的信息不对称问题,为更有效地整合不同的数据源和贷款违约预测模型,提出一种集成学习的训练方法,使用AUC(Area Under Curve)值和Q统计值对学习器的准确性和多样性进行度量,并实现了基于AUC和Q统计值的集成学习训练算法(TABAQ)。基于个人对个(P2P)贷款数据进行实证分析,发现集成学习的效果与基学习器的准确性和多样性关系密切,而与所集成的基学习器数量相关性较低,并且各种集成学习方法中统计集成表现最好。实验还发现,通过融合借款人端和投资人端的信息,可以有效地降低贷款违约预测中的信息不对称性。TABAQ能有效发挥数据源融合和学习器集成两方面的优势,在保持预测准确性稳步提升的同时,预测的一类错误数量更是进一步下降了4.85%。展开更多
文摘Although high-dose methotrexate(HD-MTX)is the most effective drug against primary CNS lymphomas(PCNSL),outcome-de-termining variables related to its administration schedule have not been defined.The impact on toxicity and outcome of the area under thecurve(AUC(MTX)),dose intensity(DI(MTX))and infusion rate(IR(MTX))of MTX and plsamatic creatinine clearance(CL(crea))was investigated in a retrospective series of 45 PCNSL patients treated with three different HD-MTX-basedcombinations.Anticon-vulsants were administered in 31 pts(69%).Age>60 years,anticonvulsant therapy,slow IR(MTX)(</=800 mgm(-2)h(-1)),and reduced DI(MTX)(</=1000 mgm(-2)wk(-1))were significantly correlated with low AUC(MTX)values.Seven pa-
文摘In diagnostic trials, clustered data are obtained when several subunits of the same patient are observed. Within-cluster correlations need to be taken into account when analyzing such clustered data. A nonparametric method has been proposed by Obuchowski (1997) to estimate the Receiver Operating Characteristic curve area (AUC) for such clustered data. However, Obuchowski’s estimator gives equal weight to all pairwise rankings within and between cluster. In this paper, we modify Obuchowski’s estimate by allowing weights for the pairwise rankings vary across clusters. We consider the optimal weights for estimating one AUC as well as two AUCs’ difference. Our results in this paper show that the optimal weights depends on not only the within-patient correlation but also the proportion of patients that have both unaffected and affected units. More importantly, we show that the loss of efficiency using equal weight instead of our optimal weights can be severe when there is a large within-cluster correlation and the proportion of patients that have both unaffected and affected units is small.
文摘接收者操作特性(Receiver operating characteristics,ROC)曲线下面积(Area under the ROC curve,AUC)常被用于度量分类器在整个类先验分布上的总体分类性能.原始Boosting算法优化分类精度,但在AUC度量下并非最优.提出了一种AUC优化Boosting改进算法,通过在原始Boosting迭代中引入数据重平衡操作,实现弱学习算法优化目标从精度向AUC的迁移.实验结果表明,较之原始Boosting算法,新算法在AUC度量下能获得更好性能.
文摘针对借贷过程中的信息不对称问题,为更有效地整合不同的数据源和贷款违约预测模型,提出一种集成学习的训练方法,使用AUC(Area Under Curve)值和Q统计值对学习器的准确性和多样性进行度量,并实现了基于AUC和Q统计值的集成学习训练算法(TABAQ)。基于个人对个(P2P)贷款数据进行实证分析,发现集成学习的效果与基学习器的准确性和多样性关系密切,而与所集成的基学习器数量相关性较低,并且各种集成学习方法中统计集成表现最好。实验还发现,通过融合借款人端和投资人端的信息,可以有效地降低贷款违约预测中的信息不对称性。TABAQ能有效发挥数据源融合和学习器集成两方面的优势,在保持预测准确性稳步提升的同时,预测的一类错误数量更是进一步下降了4.85%。
