In the assessment of car insurance claims,the claim rate for car insurance presents a highly skewed probability distribution,which is typically modeled using Tweedie distribution.The traditional approach to obtaining ...In the assessment of car insurance claims,the claim rate for car insurance presents a highly skewed probability distribution,which is typically modeled using Tweedie distribution.The traditional approach to obtaining the Tweedie regression model involves training on a centralized dataset,when the data is provided by multiple parties,training a privacy-preserving Tweedie regression model without exchanging raw data becomes a challenge.To address this issue,this study introduces a novel vertical federated learning-based Tweedie regression algorithm for multi-party auto insurance rate setting in data silos.The algorithm can keep sensitive data locally and uses privacy-preserving techniques to achieve intersection operations between the two parties holding the data.After determining which entities are shared,the participants train the model locally using the shared entity data to obtain the local generalized linear model intermediate parameters.The homomorphic encryption algorithms are introduced to interact with and update the model intermediate parameters to collaboratively complete the joint training of the car insurance rate-setting model.Performance tests on two publicly available datasets show that the proposed federated Tweedie regression algorithm can effectively generate Tweedie regression models that leverage the value of data fromboth partieswithout exchanging data.The assessment results of the scheme approach those of the Tweedie regressionmodel learned fromcentralized data,and outperformthe Tweedie regressionmodel learned independently by a single party.展开更多
The Liwu stratiform copper deposit is located in the northwestern Jianglang dome,western China.Current studies mainly focus on the genetic type and mineralization of this deposit.Detailed fluid inclusion characteristi...The Liwu stratiform copper deposit is located in the northwestern Jianglang dome,western China.Current studies mainly focus on the genetic type and mineralization of this deposit.Detailed fluid inclusion characteristics of metallogenic period quartz veins were studied to reveal the ore-forming fluid features.Laser Raman analysis indicates that the ore-forming fluids is a H_(2)O-NaCl-CH_(4)(-CO_(2))system.Fluid inclusions microthermometry shows a homogenization temperature of 181-375°C and a salinity of 5.26%-16.99%for the disseminated-banded Cu-Zn mineralization;but a homogenization temperature of 142-343°C and a salinity of 5.41%-21.19%for the massive-veined Cu-Zn mineralization.These features suggest a medium-high temperature and a medium salinity for the ore-forming fluids.H-O isotopic data indicates that the ore-forming fluids were mainly from the metamorphic and magmatic water,plus minor formation water.And sulfur isotopic data indicates that sulfur was mainly derived from the formation and magmatic rocks.Metallogenesis of the disseminated-banded mineralization was mainly correlated with fluid mixing and water-rock reaction;whereas that of the massive-veined mineralization was mainly correlated with fluid boiling.The genetic type of the deposit is a medium-high temperature hydrothermal deposit related to magmatism and controlled by shear zones.This study is beneficial to understand the stratiform copper deposit.展开更多
目的:基于佛罗里达大学质子放疗中心(University of Florida Health Proton Therapy Institute,UFHPTI)质子加速器在笔形束扫描模式下的临床实验数据,在DeepPlan中构建相应模型,验证模型构建的准确性并初步应用于临床前列腺癌的剂量计...目的:基于佛罗里达大学质子放疗中心(University of Florida Health Proton Therapy Institute,UFHPTI)质子加速器在笔形束扫描模式下的临床实验数据,在DeepPlan中构建相应模型,验证模型构建的准确性并初步应用于临床前列腺癌的剂量计算。方法:在DeepPlan质子模块中建立UFHPTI质子加速器的笔形束计算模型,并将剂量计算结果与临床实验数据进行对比,包括30组积分深度剂量(Integrated Depth Dose,IDD)、30组空气中质子束斑发散大小、1组多能量多点照射下的纵向扩展布拉格峰(Spread Out Bragg Peak,SOBP)和横向剂量分布,以此验证模型构建的准确性。最后以UFHPTI的两个前列腺癌临床放疗计划为指导,将DeepPlan计算结果与商用放疗计划系统RayStation计算结果通过PTW公司的VeriSoft软件进行gamma分析。结果:DeepPlan质子模块计算产生的30组IDD与UFHPTI加速器的临床实验数据平均相对误差为0.01%,最大相对误差为0.23%;30组空气质子束斑发散大小与临床实验数据平均相对误差为0.15%,最大相对误差为1.14%。在多能量多点照射下,DeepPlan质子模块计算产生的SOBP与临床实验数据平均相对误差为1.07%,最大相对误差为3.91%;横向剂量分布和临床实验数据平均相对误差为1.92%,最大相对误差为4.09%。针对两个前列腺癌的放疗计划,DeepPlan质子模块与RayStation计算的三维剂量结果在以3 mm/3%的标准下每个子野的gamma通过率都达到95%以上,其中病例1两个子野(270°和90°方向)的gamma通过率分别为96.4%和97.5%,病例2两个子野(270°和90°方向)的gamma通过率分别为99.3%和98.9%。结论:在DeepPlan中构建了与UFHPTI质子加速器相匹配的笔形束模型,该模型可初步应用于临床前列腺癌的剂量计算。展开更多
AIM:To investigate the role of subjective global assessment(SGA) in nutritional assessment and outcome prediction of Chinese patients with gastro-intestinal cancer.METHODS:A total of 751 patients diagnosed with gastro...AIM:To investigate the role of subjective global assessment(SGA) in nutritional assessment and outcome prediction of Chinese patients with gastro-intestinal cancer.METHODS:A total of 751 patients diagnosed with gastrointestinal cancer between August 2004 and August 2006 were enrolled in this study.Within 72 h after admission, SGA, anthropometric parameters, and laboratory tests were used to assess the nutritional status of each patient.The outcome variables including hospital stay, complications, and in-hospital medical expenditure were also obtained.RESULTS:Based on the results of SGA, 389(51.8%), 332(44.2%), and 30(4.0%) patients were classified into well nourished group(SGA-A), mildly to moderately malnourished group(SGA-B), and severely malnourished group(SGA-C), respectively.The prevalence of malnutrition classif ied by SGA, triceps skinfold thickness(TSF), mid-upper arm muscle circumference(MAMC), albumin(ALB), prealbumin(PA), and body mass index(BMI) was 48.2%, 39.4%, 37.7%, 31.3%, 21.7%, and 9.6%, respectively.In addition, ANOVA tests revealed significant differences in body mass index(BMI), TSF, PA, and ALB of patients in different SGA groups.The more severely malnourished the patient was, the lower the levels of BMI, TSF, PA, and ALB were(P < 0.05).χ2 tests showed a significant difference in SGA classif ication between patients receiving different types of treatment(surgery vs chemotherapy/radiotherapy).As the nutritional status classif ied by SGA deteriorated, the patients stayed longer in hospital and their medical expenditures increased significantly.Furthermore, multiple regression analysis showed that SGA and serum ALB could help predict the medical expenditures and hospital stay of patients undergoing surgery.The occurrence of complications increased in parallel with the increasing grade of SGA, and was the highest in the SGA-C group(23.3%) and the lowest in the SGA-A group(16.8%).CONCLUSION:SGA is a reliable assessment tool and helps to predict the hospital stay and medical expenditures of Chinese surgical gastrointestinal cancer patients.展开更多
Traditionally,fi sh carcass is headed and gutted to prepare a skin-on fi llet in surimi production.In this study,tail and/or belly of grass carp carcass were further cut off in the fi lleting step.Yield,composition,ge...Traditionally,fi sh carcass is headed and gutted to prepare a skin-on fi llet in surimi production.In this study,tail and/or belly of grass carp carcass were further cut off in the fi lleting step.Yield,composition,gelling properties and aroma profile of surimi as affected by filleting methods were investigated by SDS-PAGE,ELISA,SEM,magnetic resonance imaging(MRI)and electron nose.Cutting tail off increased surimi yield by 14.0%,whereas cutting belly off decreased it by 11.2%.Cutting tail or belly off signifi cantly decreased fat content,but did not change protein content,ash content,cathepsin(B,H and L)contents,nor protein patterns.Both breaking force and whiteness of surimi gel signifi cantly increased after cutting tail off.Cutting belly off slightly increased whiteness of surimi gel.The microstructure of all the surimi gels was compact and uniform,with fractal dimensions(Df)ranging from 2.81 to 2.85.As for the macrostructure,cutting tail off apparently improved the integrity of surimi gel while cutting belly off did not.Aroma profi les of the surimi prepared under different fi lleting methods could be clearly distinguished by linear discriminant analysis(LDA).Our results indicate that cutting tail off contribute positively to surimi production.展开更多
Background： In farm animals, mitochondrial DNA（mtDNA） effect on economic performance remains hot-topic for breeding and genetic selection. Here, 53 maternal lineages of Small-tailed Han sheep were used to investiga...Background： In farm animals, mitochondrial DNA（mtDNA） effect on economic performance remains hot-topic for breeding and genetic selection. Here, 53 maternal lineages of Small-tailed Han sheep were used to investigate the association of mitochondrial DNA variations and the lambing litter size.Results： Sequence sweeping of the mitochondrial coding regions discovered 31 non-synonymous mutations, and the association study revealed that T7719G in mt DNA t RNA-Lys gene was associated with litter size（P 〈 0.05）,manifesting 0.29 lambs per litter between the G and T carriers. Furthermore, using the mixed linear model, we assayed the potential association of the ovine litter size and haplogroups and multiple-level mtDNA haplotypes,including general haplotypes, assembled haplotypes of electron transport chain contained sequences（H-ETC）,mitochondrial respiratory complex contained sequences（H-MRC） and mitochondrial genes（H-gene, including polypeptide-coding genes, rRNA genes and tRNA genes）. The strategy for assembled mitochondrial haplotypes was proposed for the first time in mtDNA association analyses on economic traits, although none of the significant relations could be concluded（P 〉 0.05）. In addition, the nuclear major gene BMPR1B was significantly correlated with litter size in the flock（P 〈 0.05）, however, did not interact with mtDNA T7719G mutation（P 〉 0.05）.Conclusions： Our results highlight mutations of ovine mitochondrial coding genes, suggesting T7719G in tRNA-Lys gene be a potentially useful marker for selection of sheep litter size.展开更多
基金This research was funded by the National Natural Science Foundation of China(No.62272124)the National Key Research and Development Program of China(No.2022YFB2701401)+3 种基金Guizhou Province Science and Technology Plan Project(Grant Nos.Qiankehe Paltform Talent[2020]5017)The Research Project of Guizhou University for Talent Introduction(No.[2020]61)the Cultivation Project of Guizhou University(No.[2019]56)the Open Fund of Key Laboratory of Advanced Manufacturing Technology,Ministry of Education(GZUAMT2021KF[01]).
文摘In the assessment of car insurance claims,the claim rate for car insurance presents a highly skewed probability distribution,which is typically modeled using Tweedie distribution.The traditional approach to obtaining the Tweedie regression model involves training on a centralized dataset,when the data is provided by multiple parties,training a privacy-preserving Tweedie regression model without exchanging raw data becomes a challenge.To address this issue,this study introduces a novel vertical federated learning-based Tweedie regression algorithm for multi-party auto insurance rate setting in data silos.The algorithm can keep sensitive data locally and uses privacy-preserving techniques to achieve intersection operations between the two parties holding the data.After determining which entities are shared,the participants train the model locally using the shared entity data to obtain the local generalized linear model intermediate parameters.The homomorphic encryption algorithms are introduced to interact with and update the model intermediate parameters to collaboratively complete the joint training of the car insurance rate-setting model.Performance tests on two publicly available datasets show that the proposed federated Tweedie regression algorithm can effectively generate Tweedie regression models that leverage the value of data fromboth partieswithout exchanging data.The assessment results of the scheme approach those of the Tweedie regressionmodel learned fromcentralized data,and outperformthe Tweedie regressionmodel learned independently by a single party.
基金financially supported by National Natural Science Foundation of China(42272106,41202067)Open Fund of State Key Laboratory for Mineral Deposits Research,Nanjing University(2019-LAMD-K12)China Geological Survey(DD20211386,DD20211392,DD20179603).
文摘The Liwu stratiform copper deposit is located in the northwestern Jianglang dome,western China.Current studies mainly focus on the genetic type and mineralization of this deposit.Detailed fluid inclusion characteristics of metallogenic period quartz veins were studied to reveal the ore-forming fluid features.Laser Raman analysis indicates that the ore-forming fluids is a H_(2)O-NaCl-CH_(4)(-CO_(2))system.Fluid inclusions microthermometry shows a homogenization temperature of 181-375°C and a salinity of 5.26%-16.99%for the disseminated-banded Cu-Zn mineralization;but a homogenization temperature of 142-343°C and a salinity of 5.41%-21.19%for the massive-veined Cu-Zn mineralization.These features suggest a medium-high temperature and a medium salinity for the ore-forming fluids.H-O isotopic data indicates that the ore-forming fluids were mainly from the metamorphic and magmatic water,plus minor formation water.And sulfur isotopic data indicates that sulfur was mainly derived from the formation and magmatic rocks.Metallogenesis of the disseminated-banded mineralization was mainly correlated with fluid mixing and water-rock reaction;whereas that of the massive-veined mineralization was mainly correlated with fluid boiling.The genetic type of the deposit is a medium-high temperature hydrothermal deposit related to magmatism and controlled by shear zones.This study is beneficial to understand the stratiform copper deposit.
文摘目的:基于佛罗里达大学质子放疗中心(University of Florida Health Proton Therapy Institute,UFHPTI)质子加速器在笔形束扫描模式下的临床实验数据,在DeepPlan中构建相应模型,验证模型构建的准确性并初步应用于临床前列腺癌的剂量计算。方法:在DeepPlan质子模块中建立UFHPTI质子加速器的笔形束计算模型,并将剂量计算结果与临床实验数据进行对比,包括30组积分深度剂量(Integrated Depth Dose,IDD)、30组空气中质子束斑发散大小、1组多能量多点照射下的纵向扩展布拉格峰(Spread Out Bragg Peak,SOBP)和横向剂量分布,以此验证模型构建的准确性。最后以UFHPTI的两个前列腺癌临床放疗计划为指导,将DeepPlan计算结果与商用放疗计划系统RayStation计算结果通过PTW公司的VeriSoft软件进行gamma分析。结果:DeepPlan质子模块计算产生的30组IDD与UFHPTI加速器的临床实验数据平均相对误差为0.01%,最大相对误差为0.23%;30组空气质子束斑发散大小与临床实验数据平均相对误差为0.15%,最大相对误差为1.14%。在多能量多点照射下,DeepPlan质子模块计算产生的SOBP与临床实验数据平均相对误差为1.07%,最大相对误差为3.91%;横向剂量分布和临床实验数据平均相对误差为1.92%,最大相对误差为4.09%。针对两个前列腺癌的放疗计划,DeepPlan质子模块与RayStation计算的三维剂量结果在以3 mm/3%的标准下每个子野的gamma通过率都达到95%以上,其中病例1两个子野(270°和90°方向)的gamma通过率分别为96.4%和97.5%,病例2两个子野(270°和90°方向)的gamma通过率分别为99.3%和98.9%。结论:在DeepPlan中构建了与UFHPTI质子加速器相匹配的笔形束模型,该模型可初步应用于临床前列腺癌的剂量计算。
基金Supported by Shanghai Municipal Health Bureau Foundation No. 054049
文摘AIM:To investigate the role of subjective global assessment(SGA) in nutritional assessment and outcome prediction of Chinese patients with gastro-intestinal cancer.METHODS:A total of 751 patients diagnosed with gastrointestinal cancer between August 2004 and August 2006 were enrolled in this study.Within 72 h after admission, SGA, anthropometric parameters, and laboratory tests were used to assess the nutritional status of each patient.The outcome variables including hospital stay, complications, and in-hospital medical expenditure were also obtained.RESULTS:Based on the results of SGA, 389(51.8%), 332(44.2%), and 30(4.0%) patients were classified into well nourished group(SGA-A), mildly to moderately malnourished group(SGA-B), and severely malnourished group(SGA-C), respectively.The prevalence of malnutrition classif ied by SGA, triceps skinfold thickness(TSF), mid-upper arm muscle circumference(MAMC), albumin(ALB), prealbumin(PA), and body mass index(BMI) was 48.2%, 39.4%, 37.7%, 31.3%, 21.7%, and 9.6%, respectively.In addition, ANOVA tests revealed significant differences in body mass index(BMI), TSF, PA, and ALB of patients in different SGA groups.The more severely malnourished the patient was, the lower the levels of BMI, TSF, PA, and ALB were(P < 0.05).χ2 tests showed a significant difference in SGA classif ication between patients receiving different types of treatment(surgery vs chemotherapy/radiotherapy).As the nutritional status classif ied by SGA deteriorated, the patients stayed longer in hospital and their medical expenditures increased significantly.Furthermore, multiple regression analysis showed that SGA and serum ALB could help predict the medical expenditures and hospital stay of patients undergoing surgery.The occurrence of complications increased in parallel with the increasing grade of SGA, and was the highest in the SGA-C group(23.3%) and the lowest in the SGA-A group(16.8%).CONCLUSION:SGA is a reliable assessment tool and helps to predict the hospital stay and medical expenditures of Chinese surgical gastrointestinal cancer patients.
基金Authors gratefully acknowledge financial support from the National Key R&D Program of China(2018YFD0901003)China Agricultural Research System(CARS-45-28).
文摘Traditionally,fi sh carcass is headed and gutted to prepare a skin-on fi llet in surimi production.In this study,tail and/or belly of grass carp carcass were further cut off in the fi lleting step.Yield,composition,gelling properties and aroma profile of surimi as affected by filleting methods were investigated by SDS-PAGE,ELISA,SEM,magnetic resonance imaging(MRI)and electron nose.Cutting tail off increased surimi yield by 14.0%,whereas cutting belly off decreased it by 11.2%.Cutting tail or belly off signifi cantly decreased fat content,but did not change protein content,ash content,cathepsin(B,H and L)contents,nor protein patterns.Both breaking force and whiteness of surimi gel signifi cantly increased after cutting tail off.Cutting belly off slightly increased whiteness of surimi gel.The microstructure of all the surimi gels was compact and uniform,with fractal dimensions(Df)ranging from 2.81 to 2.85.As for the macrostructure,cutting tail off apparently improved the integrity of surimi gel while cutting belly off did not.Aroma profi les of the surimi prepared under different fi lleting methods could be clearly distinguished by linear discriminant analysis(LDA).Our results indicate that cutting tail off contribute positively to surimi production.
基金supported by the Project of Science and Technology of Hebei Province(15226308D)the National Key Basic Research Program of China(2014CB138500)
文摘Background： In farm animals, mitochondrial DNA（mtDNA） effect on economic performance remains hot-topic for breeding and genetic selection. Here, 53 maternal lineages of Small-tailed Han sheep were used to investigate the association of mitochondrial DNA variations and the lambing litter size.Results： Sequence sweeping of the mitochondrial coding regions discovered 31 non-synonymous mutations, and the association study revealed that T7719G in mt DNA t RNA-Lys gene was associated with litter size（P 〈 0.05）,manifesting 0.29 lambs per litter between the G and T carriers. Furthermore, using the mixed linear model, we assayed the potential association of the ovine litter size and haplogroups and multiple-level mtDNA haplotypes,including general haplotypes, assembled haplotypes of electron transport chain contained sequences（H-ETC）,mitochondrial respiratory complex contained sequences（H-MRC） and mitochondrial genes（H-gene, including polypeptide-coding genes, rRNA genes and tRNA genes）. The strategy for assembled mitochondrial haplotypes was proposed for the first time in mtDNA association analyses on economic traits, although none of the significant relations could be concluded（P 〉 0.05）. In addition, the nuclear major gene BMPR1B was significantly correlated with litter size in the flock（P 〈 0.05）, however, did not interact with mtDNA T7719G mutation（P 〉 0.05）.Conclusions： Our results highlight mutations of ovine mitochondrial coding genes, suggesting T7719G in tRNA-Lys gene be a potentially useful marker for selection of sheep litter size.
