Article(id=1195362267391902502, tenantId=1146029695717560320, journalId=1190317699101192196, issueId=1195362264082592240, articleNumber=1001-2494(2025)08-0866-09, orderNo=null, doi=10.11669/cpj.2025.08.011, pmid=null, cstr=null, oa=null, hot=null, price=null, onlineType=0, articleFormat=0, articleType=null, articleTypeStr=research-article, receivedDate=1728921600000, receivedDateStr=2024-10-15, revisedDate=null, revisedDateStr=null, acceptedDate=null, acceptedDateStr=null, onlineDate=1762926173303, onlineDateStr=2025-11-12, pubDate=1744646400000, pubDateStr=2025-04-15, doiRegisterDate=null, doiRegisterDateStr=null, onlineIssueDate=1762926173303, onlineIssueDateStr=2025-11-12, onlineJustAcceptDate=null, onlineJustAcceptDateStr=null, onlineFirstDate=null, onlineFirstDateStr=null, sourceXml=null, magXml=null, createTime=1762926173303, creator=13701087609, updateTime=1762926173303, updator=13701087609, issue=Issue{id=1195362264082592240, tenantId=1146029695717560320, journalId=1190317699101192196, year='2025', volume='60', issue='8', pageStart='777', pageEnd='890', issueExtLink='null', onlineDate='null', pubDate='null', beforeIssueId=null, nextIssueId=null, price=null, status=1, issueComplete=1, articleOrder=1, issueType=-1, specialIssue=0, createTime=1762926172514, creator=13701087609, updateTime=1762928092119, updator=13701087609, preIssue=null, nextIssue=null, ext={EN=IssueExt(id=1195370315556635165, tenantId=1146029695717560320, journalId=1190317699101192196, issueId=1195362264082592240, language=EN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=), CN=IssueExt(id=1195370315560829470, tenantId=1146029695717560320, journalId=1190317699101192196, issueId=1195362264082592240, language=CN, specialIssueTitle=, coverIllustrator=null, specialIssueEditor=, specialIssueAbout=)}, issueFiles=null}, startPage=866, endPage=874, ext={EN=ArticleExt(id=1195362267542897448, articleId=1195362267391902502, tenantId=1146029695717560320, journalId=1190317699101192196, language=EN, title=Development of Scoring Card Model for Distinguishing the Sources of Astragali Radix, columnId=null, journalTitle=Chinese Pharmaceutical Journal, columnName=null, runingTitle=null, highlight=null, articleAbstract=

OBJECTIVE To construct a scoring card capable of distinguishing between wild, semi-wild, and cultivated Astragali Radix. METHODS Six hundred batches of virtual data generated by TVAE deep learning were used as the training and validation sets. The training set data were binned, and the bins were adjusted and optimized to calculate the weight of evidence (WOE) for each bin. The data were encoded using WOE, and a logistic regression model was established. The model was trained on the training set and tuned using the validation set. A score of 50 was set as the threshold where the sample's positive and negative classification probabilities are equal. The baseline score and the scores corresponding to each bin were calculated using the formula and the logistic regression model equation. The sample score was determined by adding the base score and the scores corresponding to each bin for the sample, with a threshold of 50 used to judge the probability of the sample's positive or negative category. Card A and Card B were constructed to discriminate whether the Astragali Radix sample is wild or cultivated, respectively. RESULTS Sixty-four batches of real sample data were used as the test set to evaluate Card A and Card B, with classification accuracy rates of 0.86 and 0.80, respectively. CONCLUSION The scoring card can accurately discriminate the source of Astragali Radix samples, and the model is stable, reliable, easy to operate, and easy to promote.

, correspAuthors=Xianlong CHENG, Feng WEI, authorNote=null, correspAuthorsNote=null, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=null, magXml=null, pdfUrl=null, pdf=null, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=null, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=null, mapNumber=null, authorCompany=null, fund=null, authors=null, authorsList=Yan SHI, Xianlong CHENG, Feng WEI), CN=ArticleExt(id=1195362446555791485, articleId=1195362267391902502, tenantId=1146029695717560320, journalId=1190317699101192196, language=CN, title=评分卡模型用于判别黄芪不同来源的研究, columnId=1190352405612040510, journalTitle=中国药学杂志, columnName=论著, runingTitle=null, highlight=null, articleAbstract=

目的 构建能够区分野生品、仿野生品及栽培品黄芪的评分卡。方法 使用基于三元组的变分自动编码器(triplet-based variational autoencoder,TVAE)深度生成的600批虚拟数据作为训练集和验证集,对训练集数据进行分箱,并对分箱调整和优化,计算分箱的证据权重(weight of evidence,WOE)。对数据进行WOE编码,建立逻辑回归模型,使用训练集对逻辑回归模型进行训练,使用验证集对逻辑回归模型进行调参。以50分作为样品正负分类几率为1时的分数,根据公式和逻辑回归模型方程分别计算评分卡的基准分数和各分箱对应分数。评分时使用基准分数和样品所对应各分箱的分数相加,得出该样品的分数,以50分为阈值对样品正负类别的概率大小进行判定。构建了A卡和B卡分别实现黄芪样品是否为野生品和是否为栽培品的判别。结果 以64批真实样品数据作为测试集分别对A卡和B卡进行评估,A卡和B卡对测试集的分类判别准确率为0.86和0.80。结论 评分卡可以较准确地实现对黄芪样品来源的判别,模型稳定可靠,操作简单便捷且易推广。

, correspAuthors=程显隆, 魏锋, authorNote=null, correspAuthorsNote=
*魏锋,男,博士,研究员 研究方向:中药质量控制与评价 Tel:(010)53852020;
程显隆,男,博士,研究员 研究方向:中药质量控制与评价 Tel:(010)53851475
, copyrightStatement=null, copyrightOwner=null, extLink=null, articleAbsUrl=null, sourceXml=veZMFHCq5r8T0vjdIQYYWQ==, magXml=c1x/zbruxGq+xmkF1jRQlg==, pdfUrl=null, pdf=null, pdfFileSize=null, pdfExtLink=null, richHtmlUrl=null, mobilePdfUrl=null, reviewReport=null, pdfFirstPage=null, abstractGraph=FKQb++lfzSaTC7v/J28g6A==, abstractGraphContent=null, abstractVideo=null, citation=null, cebUrl=null, magXmlContent=6OgGOdE2LsHmIdR04Jr4hQ==, mapNumber=null, authorCompany=null, fund=null, authors=

石岩,男,博士,研究员 研究方向:中药质量控制与评价

, authorsList=石岩, 程显隆, 魏锋)}, authors=[Author(id=1195390815200653677, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, orderNo=0, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=0, authorType=1, ext={EN=AuthorExt(id=1195390815280345455, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, authorId=1195390815200653677, language=EN, stringName=Yan SHI, firstName=Yan, middleName=null, lastName=SHI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=National Institutes for Food and Drug Control, Beijing 102629, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1195390815343260016, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, authorId=1195390815200653677, language=CN, stringName=石岩, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=null, address=中国食品药品检定研究院, 北京 102629, bio={"content":"

石岩,男,博士,研究员 研究方向:中药质量控制与评价

"}, bioImg=null, bioContent=

石岩,男,博士,研究员 研究方向:中药质量控制与评价

, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1195390815116767593, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, xref=null, ext=[AuthorCompanyExt(id=1195390815125156202, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, companyId=1195390815116767593, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=National Institutes for Food and Drug Control, Beijing 102629, China), AuthorCompanyExt(id=1195390815133544811, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, companyId=1195390815116767593, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=中国食品药品检定研究院, 北京 102629)])]), Author(id=1195390815439729010, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, orderNo=1, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1195390815519420788, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, authorId=1195390815439729010, language=EN, stringName=Xianlong CHENG, firstName=Xianlong, middleName=null, lastName=CHENG, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, address=National Institutes for Food and Drug Control, Beijing 102629, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1195390815573946741, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, authorId=1195390815439729010, language=CN, stringName=程显隆, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, address=中国食品药品检定研究院, 北京 102629, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1195390815116767593, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, xref=null, ext=[AuthorCompanyExt(id=1195390815125156202, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, companyId=1195390815116767593, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=National Institutes for Food and Drug Control, Beijing 102629, China), AuthorCompanyExt(id=1195390815133544811, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, companyId=1195390815116767593, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=中国食品药品检定研究院, 北京 102629)])]), Author(id=1195390815653638519, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, orderNo=2, firstName=null, middleName=null, lastName=null, nameCn=null, orcid=null, stid=null, country=null, authorPic=null, dead=0, email=null, emailSecond=null, emailThird=null, correspondingAuthor=1, authorType=1, ext={EN=AuthorExt(id=1195390815724941689, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, authorId=1195390815653638519, language=EN, stringName=Feng WEI, firstName=Feng, middleName=null, lastName=WEI, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, address=National Institutes for Food and Drug Control, Beijing 102629, China, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null), CN=AuthorExt(id=1195390815779467642, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, authorId=1195390815653638519, language=CN, stringName=魏锋, firstName=null, middleName=null, lastName=null, prefix=null, suffix=null, authorComment=null, nameInitials=null, affiliation=null, department=null, xref=*, address=中国食品药品检定研究院, 北京 102629, bio=null, bioImg=null, bioContent=null, aboutCorrespAuthor=null)}, companyList=[AuthorCompany(id=1195390815116767593, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, xref=null, ext=[AuthorCompanyExt(id=1195390815125156202, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, companyId=1195390815116767593, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=National Institutes for Food and Drug Control, Beijing 102629, China), AuthorCompanyExt(id=1195390815133544811, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, companyId=1195390815116767593, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=中国食品药品检定研究院, 北京 102629)])])], keywords=[Keyword(id=1195390815922073979, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=EN, orderNo=1, keyword=scoring card), Keyword(id=1195390815980794236, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=EN, orderNo=2, keyword=Astragali Radix), Keyword(id=1195390816052097405, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=EN, orderNo=3, keyword=flavonoid), Keyword(id=1195390816115011966, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=EN, orderNo=4, keyword=logistic regression), Keyword(id=1195390816207286655, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=EN, orderNo=5, keyword=machine learning), Keyword(id=1195390816295367040, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=EN, orderNo=6, keyword=deep generative model), Keyword(id=1195390816366670209, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=CN, orderNo=1, keyword=评分卡), Keyword(id=1195390816433779074, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=CN, orderNo=2, keyword=黄芪), Keyword(id=1195390816496693635, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=CN, orderNo=3, keyword=黄酮), Keyword(id=1195390816622522756, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=CN, orderNo=4, keyword=逻辑回归), Keyword(id=1195390816677048709, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=CN, orderNo=5, keyword=机器学习), Keyword(id=1195390816748351878, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=CN, orderNo=6, keyword=深度生成模型)], refs=[Reference(id=1195390817926951317, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, doi=null, pmid=null, pmcid=null, year=2017, volume=36, issue=5, pageStart=149, pageEnd=159, url=null, language=null, rfNumber=[1], rfOrder=0, authorNames=DENG D S, ZHAO Y L, journalName=Review of Investment Studies(投资研究), refType=null, unstructuredReference=DENG D S, ZHAO Y L. A study on the construction and validation of a scoring card for micro and small enterprises' applications in commercial banks in China[J]. Review of Investment Studies(投资研究), 2017, 36(5):149-159., articleTitle=A study on the construction and validation of a scoring card for micro and small enterprises' applications in commercial banks in China, refAbstract=null), Reference(id=1195390817989865878, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, doi=null, pmid=null, pmcid=null, year=2023, volume=41, issue=6, pageStart=556, pageEnd=561, url=null, language=null, rfNumber=[2], rfOrder=1, authorNames=SHAO Y Y, ZHANG L Y, journalName=J Shenyang Normal Univ Nat Sci Ed(沈阳师范大学学报自然科学版), refType=null, unstructuredReference=SHAO Y Y, ZHANG L Y. Financial data anomaly risk scorecard model of listed manufacturing companies[J]. J Shenyang Normal Univ Nat Sci Ed(沈阳师范大学学报自然科学版), 2023, 41(6):556-561., articleTitle=Financial data anomaly risk scorecard model of listed manufacturing companies, refAbstract=null), Reference(id=1195390818048586135, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, doi=null, pmid=null, pmcid=null, year=2020, volume=null, issue=null, pageStart=null, pageEnd=null, url=null, language=null, rfNumber=[3], rfOrder=2, authorNames=null, journalName=null, refType=null, unstructuredReference=Ch.P(2020) Vol Ⅰ(中国药典2020年版.一部)[S]. 2020: 315., articleTitle=null, refAbstract=null), Reference(id=1195390818103112088, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, doi=null, pmid=null, pmcid=null, year=2024, volume=46, issue=1, pageStart=198, pageEnd=203, url=null, language=null, rfNumber=[4], rfOrder=3, authorNames=CHENG X W, GAO P, journalName=Chin Tradit Pat Med, refType=null, unstructuredReference=CHENG X W, GAO P. Research progress on the pharmacological effects of Astragali Radix based on the regulation of autophagy[J]. Chin Tradit Pat Med (中成药), 2024, 46(1):198-203., articleTitle=Research progress on the pharmacological effects of Astragali Radix based on the regulation of autophagy, refAbstract=null), Reference(id=1195390818157638041, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, doi=null, pmid=null, pmcid=null, year=2024, volume=49, issue=2, pageStart=294, pageEnd=303, url=null, language=null, rfNumber=[5], rfOrder=4, authorNames=MA D C, WANG Z, WANG M M, journalName=China J Chin Mater Med, refType=null, unstructuredReference=MA D C, WANG Z, WANG M M, et al. Research progress in Astragali Radix and its active ingredients in treatment of lung cancer[J]. China J Chin Mater Med (中国中药杂志), 2024, 49(2):294-303., articleTitle=Research progress in Astragali Radix and its active ingredients in treatment of lung cancer, refAbstract=null), Reference(id=1195390818212163994, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, doi=null, pmid=null, pmcid=null, year=2022, volume=24, issue=11, pageStart=4253, pageEnd=4259, url=null, language=null, rfNumber=[6], rfOrder=5, authorNames=CHEN J J, LI A P, ZHANG X Q, journalName=Mod Tradit Chin Med Mater Med-World Sci Technol, refType=null, unstructuredReference=CHEN J J, LI A P, ZHANG X Q. Study on the composition-activity relationship of Astragali Radix based on the combination of multiple composition-activity models[J]. Mod Tradit Chin Med Mater Med-World Sci Technol (世界科学技术-中医药现代化), 2022, 24(11):4253-4259., articleTitle=Study on the composition-activity relationship of Astragali Radix based on the combination of multiple composition-activity models, refAbstract=null), Reference(id=1195390818270884251, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, doi=null, pmid=null, pmcid=null, year=2023, volume=46, issue=4, pageStart=917, pageEnd=924, url=null, language=null, rfNumber=[7], rfOrder=6, authorNames=WANG Z, ZHANG J L, JIAO H J, journalName=Drug Eval Res, refType=null, unstructuredReference=WANG Z, ZHANG J L, JIAO H J, et al. Research progress on pharmacological action and quality control of effective components of Astragali Radix[J]. Drug Eval Res (药物评价研究), 2023, 46(4):917-924., articleTitle=Research progress on pharmacological action and quality control of effective components of Astragali Radix, refAbstract=null), Reference(id=1195390818337993116, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, doi=null, pmid=null, pmcid=null, year=2010, volume=35, issue=9, pageStart=1112, pageEnd=1115, url=null, language=null, rfNumber=[8], rfOrder=7, authorNames=YU K Z, LIU J, HONG H, journalName=China J Chin Mater Med, refType=null, unstructuredReference=YU K Z, LIU J, HONG H, et al. Ecological environment of cultivated Astragali Radix and market specification of prepared slices[J]. China J Chin Mater Med (中国中药杂志), 2010, 35(9):1112-1115., articleTitle=Ecological environment of cultivated Astragali Radix and market specification of prepared slices, refAbstract=null), Reference(id=1195390818409296285, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, doi=null, pmid=null, pmcid=null, year=2015, volume=38, issue=12, pageStart=2487, pageEnd=2492, url=null, language=null, rfNumber=[9], rfOrder=8, authorNames=ZHANG H T, WANG J, CHENG M E, journalName=null, refType=null, unstructuredReference=ZHANG H T, WANG J, CHENG M E, et al. Investigation on main cultivation of Astragalus membranaceus var. mongholicus and commercial specifications grades[J]. J Chin Med Mater(中药材), 2015, 38(12):2487-2492., articleTitle=Investigation on main cultivation of Astragalus membranaceus var. mongholicus and commercial specifications grades, refAbstract=null), Reference(id=1195390818488988062, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, doi=null, pmid=null, pmcid=null, year=2017, volume=42, issue=9, pageStart=1646, pageEnd=1651, url=null, language=null, rfNumber=[10], rfOrder=9, authorNames=PENG H S, ZHANG H T, PENG D Y, journalName=China J Chin Mater Med, refType=null, unstructuredReference=PENG H S, ZHANG H T, PENG D Y, et al. Evolution and characteristics of system, assessing quality by distinguishing features of traditional Chinese medicinal materials, of Dao-di herbs of Astragali Radix[J]. China J Chin Mater Med (中国中药杂志), 2017, 42(9):1646-1651., articleTitle=Evolution and characteristics of system, assessing quality by distinguishing features of traditional Chinese medicinal materials, of Dao-di herbs of Astragali Radix, refAbstract=null), Reference(id=1195390818547708319, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, doi=null, pmid=null, pmcid=null, year=2020, volume=45, issue=13, pageStart=3183, pageEnd=3190, url=null, language=null, rfNumber=[11], rfOrder=10, authorNames=ZHAO C G, LI C Y, YANG S, journalName=null, refType=null, unstructuredReference=ZHAO C G, LI C Y, YANG S, et al. Analysis of quality difference based on Astragalus membranaceus var. mongholicus in genuine region[J]. China J Chin Mater Med (中国中药杂志), 2020, 45(13): 3183-3190., articleTitle=Analysis of quality difference based on Astragalus membranaceus var. mongholicus in genuine region, refAbstract=null), Reference(id=1195390818631594400, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, doi=null, pmid=null, pmcid=null, year=2023, volume=54, issue=20, pageStart=6819, pageEnd=6826, url=null, language=null, rfNumber=[12], rfOrder=11, authorNames=ZHAO J Y, TIAN W C, FENG Z W, journalName=Chin Tradit Herb Drugs(中草药), refType=null, unstructuredReference=ZHAO J Y, TIAN W C, FENG Z W, et al. Effects of growth year, thickness and diameter class on content of active compounds of Astragali Radix decoction pieces[J]. Chin Tradit Herb Drugs(中草药), 2023, 54(20):6819-6826., articleTitle=Effects of growth year, thickness and diameter class on content of active compounds of Astragali Radix decoction pieces, refAbstract=null), Reference(id=1195390818774200737, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, doi=null, pmid=null, pmcid=null, year=2023, volume=46, issue=12, pageStart=3121, pageEnd=3126, url=null, language=null, rfNumber=[13], rfOrder=12, authorNames=CAO X E, LI X Y, AO Q, journalName=J Chin Med Mater(中药材), refType=null, unstructuredReference=CAO X E, LI X Y, AO Q, et al. Based on the JAK2/STAT3/VEGFA signaling pathway, explore the mechanism of action of total flavonoids from Astragali Radix on fibroblast-like synovial cells in human rheumatoid arthritis[J]. J Chin Med Mater(中药材), 2023, 46(12):3121-3126., articleTitle=Based on the JAK2/STAT3/VEGFA signaling pathway, explore the mechanism of action of total flavonoids from Astragali Radix on fibroblast-like synovial cells in human rheumatoid arthritis, refAbstract=null), Reference(id=1195390818879058338, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, doi=null, pmid=null, pmcid=null, year=2023, volume=37, issue=10, pageStart=1180, pageEnd=1192, url=null, language=null, rfNumber=[14], rfOrder=13, authorNames=GU H Y, LIU J, MA S C, journalName=Chin Pharm Aff, refType=null, unstructuredReference=GU H Y, LIU J, MA S C, et al. Research progress of Astragali Radix and prediction analysis of its quality markers[J]. Chin Pharm Aff (中国药事), 2023, 37(10):1180-1192., articleTitle=Research progress of Astragali Radix and prediction analysis of its quality markers, refAbstract=null), Reference(id=1195390818954555811, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, doi=null, pmid=null, pmcid=null, year=2023, volume=37, issue=3, pageStart=39, pageEnd=45, url=null, language=null, rfNumber=[15], rfOrder=14, authorNames=GONG J T, LI L, CONG Y, journalName=Res Pract Chin Med, refType=null, unstructuredReference=GONG J T, LI L, CONG Y, et al. Origin discrimination and quality evaluation of Astragali Radix based on mineral elements and active components[J]. Res Pract Chin Med (现代中药研究与实践), 2023, 37(3):39-45., articleTitle=Origin discrimination and quality evaluation of Astragali Radix based on mineral elements and active components, refAbstract=null), Reference(id=1195390819013276068, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, doi=null, pmid=null, pmcid=null, year=2022, volume=42, issue=7, pageStart=1120, pageEnd=1127, url=null, language=null, rfNumber=[16], rfOrder=15, authorNames=SHI Y, JIA T Y, LI X R, journalName=Chin J Pharm Anal(药物分析杂志), refType=null, unstructuredReference=SHI Y, JIA T Y, LI X R, et al. Quantification of flavonoid compounds in Astragali Radix[J]. Chin J Pharm Anal(药物分析杂志), 2022, 42(7):1120-1127., articleTitle=Quantification of flavonoid compounds in Astragali Radix, refAbstract=null), Reference(id=1195390819080384933, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, doi=null, pmid=null, pmcid=null, year=null, volume=null, issue=null, pageStart=null, pageEnd=null, url=https://arxiv.org/pdf/1802.04403, language=null, rfNumber=[17], rfOrder=16, authorNames=ISHFAQ H, HOOGI A, RUBIN D, journalName=arXiv preprint, arXiv: 1802. 04403, 2018, refType=null, unstructuredReference=ISHFAQ H, HOOGI A, RUBIN D. TVAE: Triplet-based variational autoencoder using metric learning[EB/OL]. arXiv preprint, arXiv: 1802. 04403, 2018. https://arxiv.org/pdf/1802.04403., articleTitle=Triplet-based variational autoencoder using metric learning, refAbstract=null), Reference(id=1195390819176853926, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, doi=null, pmid=null, pmcid=null, year=2024, volume=44, issue=5, pageStart=866, pageEnd=873, url=null, language=null, rfNumber=[18], rfOrder=17, authorNames=SHI Y, LI N, WEI F, journalName=Chin J Pharm Anal, refType=null, unstructuredReference=SHI Y, LI N, WEI F, et al. Research on the application of machine learning related techniques in the classification of Astragali Radix characterized by flavonoids[J]. Chin J Pharm Anal (药物分析杂志), 2024, 44(5):866-873., articleTitle=Research on the application of machine learning related techniques in the classification of Astragali Radix characterized by flavonoids, refAbstract=null)], funds=null, companyList=[AuthorCompany(id=1195390815116767593, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, xref=null, ext=[AuthorCompanyExt(id=1195390815125156202, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, companyId=1195390815116767593, language=EN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=National Institutes for Food and Drug Control, Beijing 102629, China), AuthorCompanyExt(id=1195390815133544811, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, companyId=1195390815116767593, language=CN, country=null, province=null, city=null, postcode=null, companyName=null, departmentName=null, remark=中国食品药品检定研究院, 北京 102629)])], figs=[ArticleFig(id=1195390816911929735, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=EN, label=Fig.1, caption=Basic process of the construction for scoring card system, figureFileSmall=ClsuJn94SxWrDT7K5ZWvAA==, figureFileBig=8xLxlu+1Wy2u+M1SIrzZcA==, tableContent=null), ArticleFig(id=1195390816991621512, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=CN, label=图1, caption=黄芪来源的黄酮类成分评分卡系统构建的基本流程, figureFileSmall=ClsuJn94SxWrDT7K5ZWvAA==, figureFileBig=8xLxlu+1Wy2u+M1SIrzZcA==, tableContent=null), ArticleFig(id=1195390817075507593, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=EN, label=Fig.2, caption=Scatter plot and kernel density estimation of the genuine data distribution of 64 batches of Astragali Radix

ZP-cultivated sample; FY-semi-wild sample; YS-wild sample.

, figureFileSmall=BKeWWdtbtvSePz0c6wVsZA==, figureFileBig=GdBh8AyqIYbnjEO4UDifdg==, tableContent=null), ArticleFig(id=1195390817134227850, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=CN, label=图2, caption=64批黄芪真实数据分布散点图及核密度估计图

ZP-栽培品;FY-仿野生品;YS-野生品。

, figureFileSmall=BKeWWdtbtvSePz0c6wVsZA==, figureFileBig=GdBh8AyqIYbnjEO4UDifdg==, tableContent=null), ArticleFig(id=1195390817197142411, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=EN, label=Fig.3, caption=64 batches of Astragali Radix real data scatter plot distribution correlation visualization, figureFileSmall=dPa3sgvUYLBXNcUWhww1Og==, figureFileBig=WJufU4UldmWZXyOWxQ0uIA==, tableContent=null), ArticleFig(id=1195390817268445580, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=CN, label=图3, caption=64批黄芪真实数据散点分布相关性可视化, figureFileSmall=dPa3sgvUYLBXNcUWhww1Og==, figureFileBig=WJufU4UldmWZXyOWxQ0uIA==, tableContent=null), ArticleFig(id=1195390817318777229, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=EN, label=Fig.4, caption=Learning curve for hyperparameter tuning of a logistic regression model for card A, figureFileSmall=ZQCgVup4q3djkApz10QKWg==, figureFileBig=r949hDrs0aHpZDsxZSH3RA==, tableContent=null), ArticleFig(id=1195390817373303182, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=CN, label=图4, caption=A卡逻辑回归模型调参学习曲线, figureFileSmall=ZQCgVup4q3djkApz10QKWg==, figureFileBig=r949hDrs0aHpZDsxZSH3RA==, tableContent=null), ArticleFig(id=1195390817432023439, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=EN, label=Fig.5, caption=The ROC of the logistic regression model for the validation set and test set in card A, figureFileSmall=rVc13DfRgs2pZbZGPdCF3Q==, figureFileBig=HKH3EIiPK9e9HZsr2+4NeA==, tableContent=null), ArticleFig(id=1195390817486549392, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=CN, label=图5, caption=A卡中逻辑回归模型对验证集和测试集的受试者工作特征曲线(ROC), figureFileSmall=rVc13DfRgs2pZbZGPdCF3Q==, figureFileBig=HKH3EIiPK9e9HZsr2+4NeA==, tableContent=null), ArticleFig(id=1195390817541075345, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=EN, label=Tab.1, caption=

Flavonoids content of Astragali Radix corresponding to the score points in A score card

, figureFileSmall=null, figureFileBig=null, tableContent=
Flavonoids Binning Score
Campanulin (-inf,0.072%] 42.853
(0.072%,0.101%] 14.178
(0.101%,inf] -44.237
Onospin (-inf,0.027%] 18.904
(0.027%,0.032%] 13.030
(0.032%,inf] -11.136
Calycosin (-inf,0.095%] 0.005
(0.095%,0.131%] 0.002
(0.131%,inf] -0.010
Sum (-inf,0.180%] 40.217
(0.180%,0.230%] 20.442
(0.230%,0.274%] 5.687
(0.274%,inf] -25.534
), ArticleFig(id=1195390817650127250, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=CN, label=表1, caption=

A卡黄酮类成分含量对应分数值

, figureFileSmall=null, figureFileBig=null, tableContent=
Flavonoids Binning Score
Campanulin (-inf,0.072%] 42.853
(0.072%,0.101%] 14.178
(0.101%,inf] -44.237
Onospin (-inf,0.027%] 18.904
(0.027%,0.032%] 13.030
(0.032%,inf] -11.136
Calycosin (-inf,0.095%] 0.005
(0.095%,0.131%] 0.002
(0.131%,inf] -0.010
Sum (-inf,0.180%] 40.217
(0.180%,0.230%] 20.442
(0.230%,0.274%] 5.687
(0.274%,inf] -25.534
), ArticleFig(id=1195390817717236115, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=EN, label=Tab.2, caption=

Flavonoids content of Astragali Radix corresponding to the score points in B score card

, figureFileSmall=null, figureFileBig=null, tableContent=
Flavonoids Binning Score
Campanulin (-inf,0.079%] 37.310
(0.079%,0.184%] 14.780
(0.184%,inf] -20.310
Onospin (-inf,0.039%] 14.270
(0.039%,0.102%] 5.170
(0.102%,inf] 8.680
Calycosin (-inf,0.131%] 58.880
(0.131%,0.160%] 4.470
(0.160%,0.308%] -43.340
(0.308%,inf] 1.900
Sum (-inf,0.332%] 15.670
(0.332%,0.544%] -2.160
(0.544%,inf] -7.840
), ArticleFig(id=1195390817788539284, tenantId=1146029695717560320, journalId=1190317699101192196, articleId=1195362267391902502, language=CN, label=表2, caption=

B卡黄酮类成分含量对应分数值

, figureFileSmall=null, figureFileBig=null, tableContent=
Flavonoids Binning Score
Campanulin (-inf,0.079%] 37.310
(0.079%,0.184%] 14.780
(0.184%,inf] -20.310
Onospin (-inf,0.039%] 14.270
(0.039%,0.102%] 5.170
(0.102%,inf] 8.680
Calycosin (-inf,0.131%] 58.880
(0.131%,0.160%] 4.470
(0.160%,0.308%] -43.340
(0.308%,inf] 1.900
Sum (-inf,0.332%] 15.670
(0.332%,0.544%] -2.160
(0.544%,inf] -7.840
)], attaches=null, journal=Journal(id=1190317596361715715, delFlag=0, nameCn=中国药学杂志, nameEn=Chinese Pharmaceutical Journal, nameHistory1=null, nameHistory2=null, issn=1001-2494, eissn=null, cn=11-2162/R, coden=null, periodic=3, language=CN, oaType=null, ccby=null, superviseOffice=null, ownerOffice=null, pubOffice=null, editorOffice=null, officeType=null, aims=null, clcCode=null, officeProv=null, officeCity=null, officeAddr=null, officeZip=null, officeEmail=null, officePhone=null, editDirector=null, officeDirector=null, officeDirectorPhone=null, officeStaffNum=null, officeEmpNum=null, coverPicUrl=hRN1R6HnoNwYkve/JRn0DA==, journalPrice=null, startedYear=null, abbrevIsoEn=null, journalRemark=null, publicationField=null, createdTime=1761723430007, updatedTime=1761735858241, createdBy=18614031015, updatedBy=13701087609, firstLetterCn=C, firstLetterEn=C, subjectCode=Life Sciences, subjectName=Life Sciences, subjectCodeEn=Life Sciences, subjectNameEn=null, picCn=hRN1R6HnoNwYkve/JRn0DA==, picEn=xSRntM4yOh2wVIE2w+OjYg==, jcr=null, cjcr=null, exts=[JournalExt(id=1190369724262355196, language=CN, name=中国药学杂志, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1761735858264, updatedTime=1761735858264, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://zgyxzzauthor.manuscriptcloud.com/login, submissionEditorUrl=https://zgyxzzeditor.manuscriptcloud.com/login, submissionReviewUrl=https://zgyxzzauthor.manuscriptcloud.com/login, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""}), JournalExt(id=1190369724358824189, language=EN, name=Chinese Pharmaceutical Journal, nameHistory1=null, nameHistory2=null, managedBy=, sponsoredBy=, publishedBy=, editorOffice=, officeProv=null, officeCity=null, officeAddr=, officeZip=, editDirector=, officeDirector=null, officePhone=null, coverPicUrl=null, journalRemark=, submitArticleUrl=null, websiteUrl=, createdTime=1761735858287, updatedTime=1761735858287, createdBy=13701087609, updatedBy=13701087609, submissionGuidelinesUrl=, submissionAuthorUrl=https://zgyxzzauthor.manuscriptcloud.com/login, submissionEditorUrl=https://zgyxzzeditor.manuscriptcloud.com/login, submissionReviewUrl=https://zgyxzzauthor.manuscriptcloud.com/login, submissionCeEditorUrl=, submissionAeEditorUrl=, option={"copyright":""})], databaseList=null, tenantJournalId=1190317699101192196, websiteList=[Website(id=1190317834875011552, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1190317699101192196, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/zgyxzz/CN, language=CN, createTime=1761723486870, createBy=18614031015, updateTime=1761723510130, updateBy=18614031015, name=中国药学杂志-中文, tplId=1146099689490845704, title=中国药学杂志, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1190318144041353703, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1190317834875011552, code=articleTextType, value=kx, createTime=1761723560581, updateTime=1761723560581, creator=18614031015, updator=18614031015), WebsiteProps(id=1190318144016187876, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1190317834875011552, code=banner, value=null, createTime=1761723560575, updateTime=1761723560575, creator=18614031015, updator=18614031015), WebsiteProps(id=1190318144062325226, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1190317834875011552, code=grayFlag, value=0, createTime=1761723560586, updateTime=1761723560586, creator=18614031015, updator=18614031015), WebsiteProps(id=1190318144007799267, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1190317834875011552, code=logo, value=https://castjournals.cast.org.cn/joweb/zgyxzz/CN/file/pic?fileId=puyAm9wIHqZks7K8hj8APQ==, createTime=1761723560573, updateTime=1761723560573, creator=18614031015, updator=18614031015), WebsiteProps(id=1190318144074908140, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1190317834875011552, code=minRunFlag, value=0, createTime=1761723560589, updateTime=1761723560589, creator=18614031015, updator=18614031015), WebsiteProps(id=1190318144032965094, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1190317834875011552, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/zgyxzz/CN/file/pic, createTime=1761723560579, updateTime=1761723560579, creator=18614031015, updator=18614031015), WebsiteProps(id=1190318144070713835, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1190317834875011552, code=silenceFlag, value=0, createTime=1761723560588, updateTime=1761723560588, creator=18614031015, updator=18614031015), WebsiteProps(id=1190318144024576485, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1190317834875011552, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_cn_619/, createTime=1761723560577, updateTime=1761723560577, creator=18614031015, updator=18614031015), WebsiteProps(id=1190318144049742312, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1190317834875011552, code=themeColor, value=null, createTime=1761723560583, updateTime=1761723560583, creator=18614031015, updator=18614031015), WebsiteProps(id=1190318144053936617, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1190317834875011552, code=themeStyle, value=null, createTime=1761723560584, updateTime=1761723560584, creator=18614031015, updator=18614031015)]), Website(id=1190317834937926113, webName=null, webTitle=null, webDomain=null, webCopyrigh=null, webIpcNo=null, seoTitle=null, seoKeywords=null, seoDescription=null, tenantJournalId=null, journalId=1190317699101192196, journalNameCn=null, journalNameEn=null, grayFlag=null, tenantId=1146029695717560320, platformId=null, journalGroupId=null, journalGroupNameCn=null, journalGroupNameEn=null, type=1, domain=https://castjournals.cast.org.cn/joweb/zgyxzz/EN, language=EN, createTime=1761723486885, createBy=18614031015, updateTime=1761723527689, updateBy=18614031015, name=中国药学杂志-英文, tplId=1146101810881728533, title=Chinese Pharmaceutical Journal, delFlag=0, indexPage=/home, props=[WebsiteProps(id=1190318170478051825, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1190317834937926113, code=articleTextType, value=kx, createTime=1761723566884, updateTime=1761723566884, creator=18614031015, updator=18614031015), WebsiteProps(id=1190318170461274606, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1190317834937926113, code=banner, value=null, createTime=1761723566880, updateTime=1761723566880, creator=18614031015, updator=18614031015), WebsiteProps(id=1190318170494829044, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1190317834937926113, code=grayFlag, value=0, createTime=1761723566888, updateTime=1761723566888, creator=18614031015, updator=18614031015), WebsiteProps(id=1190318170452885997, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1190317834937926113, code=logo, value=https://castjournals.cast.org.cn/joweb/zgyxzz/EN/file/pic?fileId=puyAm9wIHqZks7K8hj8APQ==, createTime=1761723566878, updateTime=1761723566878, creator=18614031015, updator=18614031015), WebsiteProps(id=1190318170507411958, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1190317834937926113, code=minRunFlag, value=0, createTime=1761723566891, updateTime=1761723566891, creator=18614031015, updator=18614031015), WebsiteProps(id=1190318170473857520, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1190317834937926113, code=picServerUrl, value=https://castjournals.cast.org.cn/joweb/zgyxzz/EN/file/pic, createTime=1761723566883, updateTime=1761723566883, creator=18614031015, updator=18614031015), WebsiteProps(id=1190318170503217653, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1190317834937926113, code=silenceFlag, value=0, createTime=1761723566890, updateTime=1761723566890, creator=18614031015, updator=18614031015), WebsiteProps(id=1190318170465468911, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1190317834937926113, code=staticResourcePath, value=https://castjournals.cast.org.cn/joweb/cast_kjdb_en_623/, createTime=1761723566881, updateTime=1761723566881, creator=18614031015, updator=18614031015), WebsiteProps(id=1190318170482246130, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1190317834937926113, code=themeColor, value=null, createTime=1761723566885, updateTime=1761723566885, creator=18614031015, updator=18614031015), WebsiteProps(id=1190318170486440435, tenantId=1146029695717560320, journalId=null, journalGroupId=null, siteId=1190317834937926113, code=themeStyle, value=null, createTime=1761723566886, updateTime=1761723566886, creator=18614031015, updator=18614031015)])], journalTitle=中国药学杂志, weixinUrl=null, journalUrl=http://www.zgyxzz.com.cn/, iacademicId=null, status=1, seqNo=null, journalTitleEn=Chinese Pharmaceutical Journal, journalPhotoCn=hRN1R6HnoNwYkve/JRn0DA==, journalPhotoEn=xSRntM4yOh2wVIE2w+OjYg==, journalFirstLetter=C, journalRecommend=null, journalNew=null, journalCollection=null, jcrJf=null, cjcrJf=null, jcrJfStr=null, cjcrJfStr=null, submissionFirstDecision=null, sciSubjectClassification=null, casSubjectClassification=null, citeScore=null, totalCitationFrequency=null, icpCode=null, psCode=null, advertisingLicenseCode=null, copyrightInformation=null, country=null, option=, provinceCode=null, provinceName=null, collectFlag=false), detailUrlCn=https://castjournals.cast.org.cn/joweb/zgyxzz/CN/10.11669/cpj.2025.08.011, detailUrlEn=https://castjournals.cast.org.cn/joweb/zgyxzz/EN/10.11669/cpj.2025.08.011, pdfUrlCn=https://castjournals.cast.org.cn/joweb/zgyxzz/CN/PDF/10.11669/cpj.2025.08.011, pdfUrlEn=https://castjournals.cast.org.cn/joweb/zgyxzz/EN/PDF/10.11669/cpj.2025.08.011, aliStartDate=null, aliEndDate=null, collectionFlag=false, citedCount=null, citedUrl=null, reference=null)
收藏切换
评分卡模型用于判别黄芪不同来源的研究
收藏切换
PDF下载
石岩 , 程显隆 * , 魏锋 *
中国药学杂志 | 论著 2025,60(8): 866-874
收起
收藏切换
中国药学杂志 | 论著 2025, 60(8): 866-874
评分卡模型用于判别黄芪不同来源的研究
全屏
石岩, 程显隆*, 魏锋*
作者信息
  • 中国食品药品检定研究院, 北京 102629
  • 石岩,男,博士,研究员 研究方向:中药质量控制与评价

通讯作者:

*魏锋,男,博士,研究员 研究方向:中药质量控制与评价 Tel:(010)53852020;
程显隆,男,博士,研究员 研究方向:中药质量控制与评价 Tel:(010)53851475
Development of Scoring Card Model for Distinguishing the Sources of Astragali Radix
Yan SHI, Xianlong CHENG*, Feng WEI*
Affiliations
  • National Institutes for Food and Drug Control, Beijing 102629, China
出版时间: 2025-04-15 doi: 10.11669/cpj.2025.08.011
文章导航
收藏切换

目的 构建能够区分野生品、仿野生品及栽培品黄芪的评分卡。方法 使用基于三元组的变分自动编码器(triplet-based variational autoencoder,TVAE)深度生成的600批虚拟数据作为训练集和验证集,对训练集数据进行分箱,并对分箱调整和优化,计算分箱的证据权重(weight of evidence,WOE)。对数据进行WOE编码,建立逻辑回归模型,使用训练集对逻辑回归模型进行训练,使用验证集对逻辑回归模型进行调参。以50分作为样品正负分类几率为1时的分数,根据公式和逻辑回归模型方程分别计算评分卡的基准分数和各分箱对应分数。评分时使用基准分数和样品所对应各分箱的分数相加,得出该样品的分数,以50分为阈值对样品正负类别的概率大小进行判定。构建了A卡和B卡分别实现黄芪样品是否为野生品和是否为栽培品的判别。结果 以64批真实样品数据作为测试集分别对A卡和B卡进行评估,A卡和B卡对测试集的分类判别准确率为0.86和0.80。结论 评分卡可以较准确地实现对黄芪样品来源的判别,模型稳定可靠,操作简单便捷且易推广。

评分卡  /  黄芪  /  黄酮  /  逻辑回归  /  机器学习  /  深度生成模型

OBJECTIVE To construct a scoring card capable of distinguishing between wild, semi-wild, and cultivated Astragali Radix. METHODS Six hundred batches of virtual data generated by TVAE deep learning were used as the training and validation sets. The training set data were binned, and the bins were adjusted and optimized to calculate the weight of evidence (WOE) for each bin. The data were encoded using WOE, and a logistic regression model was established. The model was trained on the training set and tuned using the validation set. A score of 50 was set as the threshold where the sample's positive and negative classification probabilities are equal. The baseline score and the scores corresponding to each bin were calculated using the formula and the logistic regression model equation. The sample score was determined by adding the base score and the scores corresponding to each bin for the sample, with a threshold of 50 used to judge the probability of the sample's positive or negative category. Card A and Card B were constructed to discriminate whether the Astragali Radix sample is wild or cultivated, respectively. RESULTS Sixty-four batches of real sample data were used as the test set to evaluate Card A and Card B, with classification accuracy rates of 0.86 and 0.80, respectively. CONCLUSION The scoring card can accurately discriminate the source of Astragali Radix samples, and the model is stable, reliable, easy to operate, and easy to promote.

scoring card  /  Astragali Radix  /  flavonoid  /  logistic regression  /  machine learning  /  deep generative model
石岩, 程显隆, 魏锋. 评分卡模型用于判别黄芪不同来源的研究. 中国药学杂志, 2025 , 60 (8) : 866 -874 . DOI: 10.11669/cpj.2025.08.011
Yan SHI, Xianlong CHENG, Feng WEI. Development of Scoring Card Model for Distinguishing the Sources of Astragali Radix[J]. Chinese Pharmaceutical Journal, 2025 , 60 (8) : 866 -874 . DOI: 10.11669/cpj.2025.08.011
基于模型对药物的质量进行分析判断是目前机器学习技术在药物分析领域内使用的主要形式。在药物分析实验所获得数据的进一步处理、分析和判断过程中,机器学习技术具有快速、高效等无可比拟的优势,但是在实际使用中仍面临模型传递困难的问题,并且对使用者也有一定的机器学习相关的专业性要求,严重阻碍了机器学习技术在药物分析领域内的进一步应用。为解决这一问题,本研究提出了基于机器学习的评分卡技术作为解决方案。
评分卡是智能风控领域内极为重要且被广泛应用的一种工具,在信贷审批、信用卡审批、保险审批、资产管理及欺诈检测等多种涉及资产及金融安全的实际业务中具有重要地位,对于金融系统安全、稳定和高效地运行有至关重要的作用[1-2]。本质上,评分卡是根据所制定的特征分箱赋分规则,基于被评估对象实际特征数据对该评估对象进行评分,以获得被评估对象对于实际业务的风险估计。在评分卡系统构建过程中,须结合统计机器学习的相关知识与技术,评分卡系统构建完成后,实际使用环节丝毫不涉及专业的知识与技术,基层业务人员仅需简单操作便可完成评估工作。
黄芪是一种十分常用的中药材,为豆科植物蒙古黄芪[Astragalus membranaceus(Fisch.) Bge. var. mongholicus(Bge.)Hsiao]或膜荚黄芪[Astragalus membranaceus(Fisch.) Bge.]的干燥根,主要具有补气升阳、固表止汗等功效,在抗炎、抗氧化、抗衰老及免疫调节等方面具有一定生物活性,其中的黄酮类、皂苷类和多糖类等是主要活性成分[3-7]。由于黄芪野生资源逐步减少,目前市场上的黄芪以蒙古黄芪的栽培品为主,膜荚黄芪仅在东北部分地区有少量分布[8]。黄芪市场销量很大,栽培品集中在甘肃、山西和内蒙古等地,而在山西部分地区也分布着野生和仿野生(半野生)的黄芪,这些来源的黄芪被公认为黄芪的道地药材,质量最佳,价格也最为昂贵[8-11]。目前多项研究[12-15]结果表明,黄酮类成分是黄芪中十分重要的一类活性成分,与黄芪的传统功效具有一致性,同时也是判别不同来源黄芪的重要指标。
本课题组前期研究[16]获得大量真实实验测定黄酮类成分的数据,在此基础上,本研究使用深度学习技术中常用的基于三元组的变分自动编码器[17](triplet-based variational autoencoder,TVAE)深度生成仿真样品系列黄酮类成分数据,而这些数据已经由研究[18]证实了对真实数据的可替代性。本研究使用了TVAE技术深度生成的仿真数据,并基于评分卡特征分箱、证据权重(weight of evidence,WOE)计算、逻辑回归模型建立及分数设计等步骤构建了两套基于黄酮类成分含量的黄芪来源评分卡系统,可以分别用来判别黄芪的野生品、仿野生品及栽培品,并且使用真实实验测定的数据对评分卡的使用效果进行基于真实数据的评估,取得了满意结果,具体的研究流程见图1
本研究样品共64批,其中人工栽培(ZP)样品30批,仿野生(FY)样品24批,野生(YS)样品10批。ZP样品收集于中药饮片销售和使用单位,均为1~2年生,FY样品采集于山西应县,YS样品采集于山西浑源县。基于真实数据和TVAE技术深度生成的仿真虚拟数据共600批,其中ZP、FY和YS样品每个类别各200批,TVAE算法参数epochs和batch_size分别为400和10。真实和虚拟数据在经过训练的14种机器学习模型中预测指标趋势基本一致,彼此具有较好的相似性,可以用于进一步的人工智能模型研究。
本研究对于数据的处理、探索性数据分析(exploratory data analysis, EDA)、评分卡系统构建等步骤均使用计算机编程语言Python(美国Python Software Foundation,版本:3.8.8)编写代码并运行完成。研究中所涉及的所有随机数种子取值均为837。
真实数据由64批共6种黄酮类成分的含量组成,加上这6种黄酮类成分含量总和,可组成64×7的数据矩阵。将各批样品的数据按照各黄酮类成分及总和两两绘制散点图及核密度估计图,并以不同颜色区分样品种类,见图2
由以上散点图,进一步选取毛蕊异黄酮葡萄糖苷、刺芒柄花苷、毛蕊异黄酮、芒柄花素和总和数据进行散点相关性可视化分析,见图3
数据为3个类别,应用解决二分类问题的评分卡技术须采用构建2套评分卡的方式。根据评分卡系统的实际使用情景,本研究首先构建了判别黄芪是否为野生品的评分卡系统(A卡),即区分野生品(YS)和非野生品(FY和ZP),随后又构建了判别黄芪是否为栽培品的评分卡系统(B卡),即区分非栽培品(YS和FY)和栽培品(ZP)。在构建A卡系统时,YS样品类别编码为0,FY和ZP样品类别编码为1;构建B卡系统时,YS和FY样品类别编码为0,ZP样品类别编码为1。
将600批深度生成的数据打乱排序,然后按照8∶2比例随机划分训练集和验证集,训练集共480批,供特征变量(即本文中的各黄酮类成分含量)数据分箱、WOE数值映射计算及训练模型使用,验证集共120批,供验证特征分箱和WOE数值映射计算规则使用,并在WOE数值转换后用于逻辑回归模型超参数的优化调整。
64批真实数据作为测试集,按照所构建评分卡系统流程对测试集样品数据进行打分和判断,用于对所构建的评分卡系统的整体测试和评估。
在构建A卡和B卡系统的分箱规则时,首先将训练集数据的各特征变量分别进行等频分箱,A卡和B卡系统初始分箱个数均为20个。以每个分箱中必须同时含有编号为0和1的样品为原则进行分箱合并。若需合并,首个分箱向下与相邻分箱合并,其他分箱向上与相邻分箱合并。按照该方法,对A卡和B卡分别进行初步的分箱数据合并。
按照A卡和B卡系统分别的样品类别0和1编码及各自特征变量数据分箱情况,按照公式1~2计算这2种评分卡系统的各特征变量的各分箱的WOE及对应的IV。
WOEi=ln$\left(\frac{{N}_{i}/{N}_{T}}{{P}_{i}/{P}_{T}}\right)$
IV=$\stackrel{n}{\sum _{i=1}}$(Ni/NT-Pi/PT)ln$\left(\frac{{N}_{i}/{N}_{T}}{{P}_{i}/{P}_{T}}\right)$
公式1~2中,Ni为某一黄酮类成分的第i个分箱中0类别样品的数量,NT为数据集中0类别样本的总数,Pi为某一黄酮类成分的第i个分箱中1类别样品的数量,PT为数据集1类别样本的总数。
以各特征变量的分箱中所包含的编码分别为0和1的样品类别的样品数为数据,分别对各相邻分箱进行卡方检验并计算P值。基于P值由大到小进行相邻分箱合并的原则进行逐步地合并分箱操作,并计算每步合并后的该特征变量的IV,绘制分箱数与IV的变化折线图。以斜率最大的该段折线前后两点的分箱数为分箱合并目标,结合P值(一般均应不大于0.05)确定分箱数,分箱数一般应为3~5个。由于部分样品的芒柄花素、异鼠李素和山柰酚的含量较低,原数据中为未检出标示,深度生成的数据数值也较低,在进行分箱和合并处理时无法操作,故舍去这3种黄酮类成分的数据,但总和数据中仍包含这3种黄酮类成分的含量。A卡和B卡的分箱情况见表1~2。
确定分箱之后,按照公式1计算分别计算得出A卡和B卡各自的特征变量分箱数据段所对应的WOE,并以各批样品数据对应的分箱的WOE替代该批样品的特征变量具体数值。以经过WOE转换后的样品数据进行后续的机器学习建模等步骤。
分别建立A卡和B卡的逻辑回归模型,并以验证集对模型的超参数进行优化。A卡的逻辑回归模型的正则化惩罚项选择L1法,正则化强度倒数C为0.1,求解器为“saga”,最大迭代次数为20;B卡的逻辑回归模型的正则化惩罚项选择L2法,正则化强度倒数C为0.1,求解器为“lbfgs”,最大迭代次数为20。
若一个事件发生的概率为P,那么该事件不发生的概率为1-P,该事件发生的几率(odds)即为P/(1-P)。A卡系统是用于判别黄芪是否为野生品的评分卡,本研究设定:当样品为野生品(YS)的概率和非野生品(FY和ZP)的概率各为50%时,即样品为非野生品(FY和ZP)的几率为1时,得分为50分;当样品为非野生品(FY和ZP)的几率为翻倍为2时,翻倍分数值(point of double,PDO)设为10,即此时得分为40分。基于以上设定及公式3可列二元一次方程,经过计算可得补偿数A为50.00,刻度数B为14.43。
score=A-B×log(odds)
公式3中score为评分卡系统的得分,odds为样品为非野生品(FY和ZP)的几率。逻辑回归模型是基于对数几率的学习模型,模型类别的几率对数以拟合为线性方程形式表达,按照传统线性代数中的矩阵表达方式可记为${\widehat{\omega }}^{T}$·$\hat{X}$。当$\hat{X}$中的x为0时,ω为该方程的截距(intercept),把该截距代入公式3可推导出基础分数(base score,BS)为A-B×intercept,A卡系统逻辑回归模型中方程的截距按该法代入的计算可得A卡系统BS为47.674。将A卡系统逻辑回归模型中各特征变量的系数ω与各分箱对应的WOE数值分别相乘,再与公式3中的刻度数B相乘并取负值,便可得出A卡系统各黄酮类成分含量对应的分数值,将这一数值与BS相加,便可得到待评分黄芪样品的分数。当该分数大于50分时,表明该样品是非野生品的几率小于1,即野生品的概率大于非野生品,该分数小于50分时,表明该样品是非野生品的几率大于1,分数小于40分时表明非野生品的几率大于2。A卡系统中各黄酮类数据对应分箱及对应的分数值见表1
B卡系统是用于判别黄芪是否为栽培品的评分卡,与上述A卡系统相同的分数设定及算法可计算得到B卡系统的BS为62.716,及B卡系统中各黄酮类数据各分箱对应的分数值,分箱数据及对应的分数值见表2。B卡系统的用法与A卡系统相似,将待评分样品的黄酮类成分各含量与表2对应分箱的分数值和B卡系统的BS相加,所得即为B卡评分。同样以50分为界,大于50分的样品是野生品或半野生品(即非栽培品)的概率大于其是栽培品的概率,小于50分则表明样品为栽培品的概率大于非栽培品的概率,分数越高表明样品为非栽培品的概率增加,反之则表明样品为栽培品的概率增加。当得分为40分时,样品为栽培品的几率为2,栽培品的概率2倍于非栽培品的概率。
评分卡在使用过程中,需要根据待评分黄芪样品的毛蕊异黄酮葡萄糖苷、毛蕊异黄酮、刺芒柄花苷、芒柄花素、异鼠李素、槲皮素的含量,与评分卡的分箱对应分数(表1表2)与对应评分卡的BS相加,根据得分与50分比较大小,即可得到评分判别结果。
以上A卡和B卡的评分卡模型都是基于深度生成的仿真数据,64批真实的样品数据用于对A卡和B卡的评分效果进行验证和评估。
A卡对64批真实样品进行评分,并以50分为界,依据分数判断是否为野生品,结果64批样品评分判断正确的为55批,整体准确率为0.86。30批栽培品样品中有26批准确判断为非野生品,准确率为0.87;24批半野生品中有21批准确判断为非野生品,准确率为0.88;10批野生品样品中有8批准确判断为野生品,准确率为0.80。
B卡对64批真实样品进行评分,并按照50分为界,依据分数判断是否为栽培品,结果64批样品评分判断正确的为51批,整体准确率为0.80。30批栽培品样品中有23批准确判断为栽培品,准确率为0.77;24批半野生品中有19批准确判断为非栽培品,准确率为0.79;10批野生品样品中有9批准确判断为非栽培品,准确率为0.90。
由于真实样品数据量较少,且数据存在类别不平衡的情况。智能风控领域的数据经常也会面临类别不平衡的情况,在构建评分卡时可通过上采样等技术对数据进行处理。结合实际情况,本研究选择基于TVAE深度生成仿真数据的方式对数据进行增强及平衡化处理。研究[18]结果表明,使用TVAE深度生成的数据训练得到的14个模型对于真实样品数据的分类准确率为0.813~0.891,且这14个模型的类别涵盖了树模型、神经网络模型和k近邻模型等不同原理的机器学习模型,充分证实了TVAE深度生成数据的可靠性。而本研究中基于TVAE深度生成的数据所构建的评分卡对于真实数据的评分判断准确率结果也与14个模型的分类准确率结果具有较好的一致性。
一系列研究的结果表明,数据增强技术可以较大程度解决以真实实验数据为基础的样本数据量不足的问题,有助于使用较少数据构建人工智能模型的研究和应用。然而,为避免数据偏差造成的模型训练错误,建议使用真实实验数据对训练后的模型进行测试和评估,而且在模型部署并投入实际应用后,应该不断使用新获得的真实数据进行模型的增量学习,以不断改进增强模型。
在机器学习中,常见的用于建模的特征变量一般可分为数值型和类别型等类型,其中数值型可分为连续型和离散型。评分卡构建过程中将原始特征变量进行WOE转换的过程,本质就是WOE编码的过程,通过WOE编码可以使类别型数据变为连续型数据,相较于独热编码,WOE编码避免了数据稀疏化,更利于后续步骤中逻辑回归模型的应用。WOE编码的步骤是先对原始特征变量的数据进行分箱处理,即数据转为离散化,然后根据分箱中好坏样本的比例将离散化的数据转换为连续型的变量数据。这样处理不仅增加了特征变量的可解释性,还可以使得非线性相关的变量成为线性相关的变量,适用于缺失值变量的处理,并且可以过滤掉变量的不正常波动,显著增加模型的稳定性。此外,WOE编码后还可以减少特征变量之间的相关性,降低模型的过拟合风险。所谓好坏样本是智能风控领域的描述,归其本质而言好坏样本或者风险的低高是表现在评分卡最终得分,一般来说分数越高表明评估对象是好客户的概率越大,投资的风险越低。在本研究中,基于黄芪的真实情况,设定为分类级别高的黄芪类别为好样本,反之为坏样本,以此设定计算分箱的WOE。
一般来说,IV数值越大表明该特征变量对于模型分类的贡献越大,因此常用于对特征变量的筛选,在评分卡的构建过程中IV的应用是非常重要的环节。然而IV数值并非越大越好,因为IV会随着分箱数量的增加而增大,分箱过多会导致单个分箱中样本量的减少,会极大影响分箱中样本分布的稳定性。在本研究中,由于部分样品的芒柄花素、异鼠李素和山柰酚的含量较低,在计算IV时表现为无法计算或分箱数量增加后IV无变化,从这一方面剔除了这3个黄酮类特征。另外,结合卡方检验的P值,通过观察IV的变化趋势确定了各特征变量的分箱数量。
机器学习应用中,对模型进行调参是重要的一个环节。逻辑回归模型的调参一般会涉及正则化惩罚项选择、求解器的选择和迭代次数等。机器学习的调参须避免对模型的数据泄露,否则可能会导致过拟合等问题,而调参不到位也可能会出现模型的欠拟合,因此模型的调参应使用专门的验证集对模型参数的可能值进行尽量详尽的反复验证。借助于较多的深度生成的数据,本研究采用了随机划分的120批深度生成的样品数据作为验证集,通过绘制学习曲线的方式对逻辑回归模型进行调参。以对A卡逻辑回归模型的正则化惩罚项选择为例,图4为分别使用两种正则化方法(L1和L2)情况下,不同取值的正则化强度倒数C的学习曲线。
逻辑回归模型调参完毕,使用训练集数据对逻辑回归模型进行训练,然后再使用64批真实样品数据组成的测试集对逻辑回归模型的分类效果进行评估。A卡和B卡构建过程中的逻辑回归模型对于测试集的分类准确率分别为0.84和0.72。对于分类问题而言,仅仅依靠准确率进行评估并不能准确全面反映模型效能,因此常常引入其他的评估指标。本研究通过绘制受试者工作特征曲线(receiver operating characteristic curve,ROC)并使用曲线下面积(area under the curve,AUC)作为另一种模型评估指标。ROC的横坐标是假阳性率(false positive rate),纵坐标是真阳性率(true positive rate),根据模型对各类别实际分类情况分别进行曲线的绘制,可以简单直观地通过可视化的图像方式观察模型的分类准确性,能够准确地反映模型分类的假阳性率和真阳性率的关系,如果对其使用精确的数值评估,须分别计算各类别的AUC。AUC数值介于0~1之间,使用AUC评估模型分类效果,数值越大代表模型分类效果越好。由于本研究的A卡和B卡都是二分类问题,因此每个逻辑回归模型的2个类别的AUC是相同的,其中A卡和B卡中的逻辑回归模型对测试集分类效果的AUC分别为0.90和0.78。以分别使用验证集和测试集对A卡中的逻辑回归模型评估并绘制ROC为例,见图5
图5可知,逻辑回归模型在验证集上的表现极佳,显著强于模型在测试集上的表现,其原因首先是验证集和测试集分别源自基于真实数据深度生成的虚拟数据和真实样品数据,两者本质上不可能完全一致,其次是使用验证集对模型进行调参,再考察模型对验证集的表现,存在了数据泄露问题。因此本研究中逻辑回归模型的实际表现应以测试集数据为准。
逻辑回归模型属于广义的线性回归模型,是通过将对数几率函数作为联系函数应用在常规的线性回归方程并经过变化而得。其中的S形函数常被称为sigmoid函数,可以将输入映射在0和1之间的概率值上,一般可代表模型判断样品属于某类别的可能性,当输入为0时,sigmoid函数映射的数值为0.5,代表模型判断样品为二分类的可能性相同,即概率相同,几率为1。在这里所谓的概率,其实是机器学习模型对样品的类别判断的概率,这种概率的准确程度随着机器学习模型对样品预测能力的增强而增加。本研究在进行分数设计的阶段所进行的几率和概率的设定正是基于此原理,通过评分卡计算得到的分数也可以计算得到相应的概率。
评分分数的设计是灵活可变的,考虑药物分析领域使用习惯本研究以百分内数值为基准,将50分作为类别概率相等的界限。这种分数设计可根据需求进行调整和修改,随着分数设计的调整,各分箱对应的得分值以及BS也必然会发生变化。
评分卡的构建是基于稳健且具备良好可解释性的逻辑回归模型算法的,而机器学习模型的训练源自样品特征变量数据,其评分结果可以与特征变量相对应,这种可解释性利于评分卡评分机制的透明化,利于该技术的推广及标准化。同时,根据评分卡的这种可解释性,还可以为评价对象的进一步深入研究提供强有力的参考和佐证作用。
表1表明,毛蕊异黄酮葡萄糖苷、总和(6种黄酮)及刺芒柄花苷对于A卡的评分具有重要的作用,考虑含量数值的分布,毛蕊异黄酮葡萄糖苷和刺芒柄花苷这2种黄酮类成分对于A卡评分的影响最大。这2种黄酮类成分的含量数值偏高的分箱区间对应分数为负数,结合样品得分越高,野生品的概率越大的设定,可得出毛蕊异黄酮葡萄糖苷和刺芒柄花苷含量越高样品是野生品的概率越小结论,当毛蕊异黄酮葡萄糖苷含量≤0.072%及总和≤0.180%时,样品是野生品的概率极大地增加。同理由表2可推知,毛蕊异黄酮和毛蕊异黄酮葡萄糖苷这2种黄酮类成分对于B卡评分的影响最大,一般来讲,这2种黄酮类成分的含量越高,样品是栽培品的概率越大。如果毛蕊异黄酮含量≤0.131%且毛蕊异黄酮葡萄糖苷含量≤0.079%时,样品仅此2项得分已达96.19分,根据B卡判定必为非栽培品。如果样品毛蕊异黄酮的含量在0.160%与 0.308%之间且毛蕊异黄酮葡萄糖苷含量≥0.184%时,B卡则定会判为栽培品。需要指出的是,评分卡模型在实际使用中是不断调整改进的过程,并非一成不变,本研究所使用的数据量毕竟有限,且模型构建过程多依赖深度生成数据,随着使用中数据量的积累增加,评分卡的分箱会越来越准确合理,其可解释的数据也会越来越接近真实。
不同于常用的赋值打分的模式,评分卡的评分是基于对各特征变量数据分布区域的分箱处理方式,对于源自自然界的中药材及饮片可能更加适合。中药材及饮片个体、生长环境和加工方式等差异在所难免,其成分指标也不可能十分精确于某固定数值,而应该呈一定规律的区间分布,这点与评分卡的分箱处理十分契合,而且较好避免了由于指标数据的波动引起的结果偏差,评分结果更加稳定。此外,分箱结果的最低值与最高值分别以负无穷和正无穷代替,也避免了异常值对结果造成的大的偏差波动,进一步增加了稳定性。
机器学习模型的使用一般需要相关人员具备一定的专业知识,而且在目前药物分析领域内,电子化模型的部署和应用也存在一定的问题这些都限制了机器学习技术在药物分析领域内的发展和应用。
智能风控领域中,评分卡技术诞生的初衷就是方便无相关专业背景的工作人员仅通过对评估对象进行简单打分和求和,对照分数的说明便可以完成金融风险的评估工作,这个模式在保证评估正确率的同时,极大地提高模型的使用效率和便捷程度。本研究借鉴了评分卡技术,并与药物分析领域相关专业和知识相结合,所构建的A卡和B卡均可实现仅需通过对黄芪样品中的几种黄酮类成分含量所归属的区间(分箱)所对应的分数进行简单地评分,便可对黄芪样品进行较准确的级别分类,整个过程无需对数据进行标准化处理,也无需人员具备专业的机器学习和药物分析领域知识。
评分卡的稳定性和实用性表现极佳,加之其对于风险判定结果的可解释性,使得评分卡自诞生以来一直是智能风控领域内主要的技术应用之一。本研究参考了智能风控领域的评分卡相关技术,并与药物分析的自身实际情况相结合,创新引入深度生成数据,所构建的A卡和B卡用于黄芪饮片来源的判别,通过真实样品测试,取得较好的分类效果。本研究是数据驱动下的药品评价新模式的探索和试验,是人工智能和药物分析两个学科融合解决问题的尝试,对新机遇新形势下药物分析专业的发展具有积极的意义。
参考文献 引证文献
排序方式:
[1]
DENG D S, ZHAO Y L. A study on the construction and validation of a scoring card for micro and small enterprises' applications in commercial banks in China[J]. Review of Investment Studies(投资研究), 2017, 36(5):149-159.
[2]
SHAO Y Y, ZHANG L Y. Financial data anomaly risk scorecard model of listed manufacturing companies[J]. J Shenyang Normal Univ Nat Sci Ed(沈阳师范大学学报自然科学版), 2023, 41(6):556-561.
[3]
Ch.P(2020) Vol Ⅰ(中国药典2020年版.一部)[S]. 2020: 315.
[4]
CHENG X W, GAO P. Research progress on the pharmacological effects of Astragali Radix based on the regulation of autophagy[J]. Chin Tradit Pat Med (中成药), 2024, 46(1):198-203.
[5]
MA D C, WANG Z, WANG M M, et al. Research progress in Astragali Radix and its active ingredients in treatment of lung cancer[J]. China J Chin Mater Med (中国中药杂志), 2024, 49(2):294-303.
[6]
CHEN J J, LI A P, ZHANG X Q. Study on the composition-activity relationship of Astragali Radix based on the combination of multiple composition-activity models[J]. Mod Tradit Chin Med Mater Med-World Sci Technol (世界科学技术-中医药现代化), 2022, 24(11):4253-4259.
[7]
WANG Z, ZHANG J L, JIAO H J, et al. Research progress on pharmacological action and quality control of effective components of Astragali Radix[J]. Drug Eval Res (药物评价研究), 2023, 46(4):917-924.
[8]
YU K Z, LIU J, HONG H, et al. Ecological environment of cultivated Astragali Radix and market specification of prepared slices[J]. China J Chin Mater Med (中国中药杂志), 2010, 35(9):1112-1115.
[9]
ZHANG H T, WANG J, CHENG M E, et al. Investigation on main cultivation of Astragalus membranaceus var. mongholicus and commercial specifications grades[J]. J Chin Med Mater(中药材), 2015, 38(12):2487-2492.
[10]
PENG H S, ZHANG H T, PENG D Y, et al. Evolution and characteristics of system, assessing quality by distinguishing features of traditional Chinese medicinal materials, of Dao-di herbs of Astragali Radix[J]. China J Chin Mater Med (中国中药杂志), 2017, 42(9):1646-1651.
[11]
ZHAO C G, LI C Y, YANG S, et al. Analysis of quality difference based on Astragalus membranaceus var. mongholicus in genuine region[J]. China J Chin Mater Med (中国中药杂志), 2020, 45(13): 3183-3190.
[12]
ZHAO J Y, TIAN W C, FENG Z W, et al. Effects of growth year, thickness and diameter class on content of active compounds of Astragali Radix decoction pieces[J]. Chin Tradit Herb Drugs(中草药), 2023, 54(20):6819-6826.
[13]
CAO X E, LI X Y, AO Q, et al. Based on the JAK2/STAT3/VEGFA signaling pathway, explore the mechanism of action of total flavonoids from Astragali Radix on fibroblast-like synovial cells in human rheumatoid arthritis[J]. J Chin Med Mater(中药材), 2023, 46(12):3121-3126.
[14]
GU H Y, LIU J, MA S C, et al. Research progress of Astragali Radix and prediction analysis of its quality markers[J]. Chin Pharm Aff (中国药事), 2023, 37(10):1180-1192.
[15]
GONG J T, LI L, CONG Y, et al. Origin discrimination and quality evaluation of Astragali Radix based on mineral elements and active components[J]. Res Pract Chin Med (现代中药研究与实践), 2023, 37(3):39-45.
[16]
SHI Y, JIA T Y, LI X R, et al. Quantification of flavonoid compounds in Astragali Radix[J]. Chin J Pharm Anal(药物分析杂志), 2022, 42(7):1120-1127.
[17]
ISHFAQ H, HOOGI A, RUBIN D. TVAE: Triplet-based variational autoencoder using metric learning[EB/OL]. arXiv preprint, arXiv: 1802. 04403, 2018. https://arxiv.org/pdf/1802.04403. https://arxiv.org/pdf/1802.04403
[18]
SHI Y, LI N, WEI F, et al. Research on the application of machine learning related techniques in the classification of Astragali Radix characterized by flavonoids[J]. Chin J Pharm Anal (药物分析杂志), 2024, 44(5):866-873.
2025年第60卷第8期
PDF下载
84
33
引用本文
BibTeX
文章信息
doi: 10.11669/cpj.2025.08.011
  • 接收时间:2024-10-15
  • 首发时间:2025-11-12
  • 出版时间:2025-04-15
补充材料
相关文章
文章信息
作者
出版历史
  • 收稿日期:2024-10-15
基金
作者信息
    中国食品药品检定研究院, 北京 102629

通讯作者:

*魏锋,男,博士,研究员 研究方向:中药质量控制与评价 Tel:(010)53852020;
程显隆,男,博士,研究员 研究方向:中药质量控制与评价 Tel:(010)53851475
参考文献
分享链接
https://castjournals.cast.org.cn/joweb/zgyxzz/CN/10.11669/cpj.2025.08.011
分享至
全文二维码

扫描看全文

引用本文
BibTeX
本文的引用情况
2种不同金属材料的力学参数

Family
属数
Number of
genus
种数
Number of
species
占总种数比例
Percentage of
total species (%)

Genus
种数
Number of
species
占总种数比例
Percentage of total
species (%)
鹅膏菌科Amanitaceae 2 11 5.26 鹅膏菌属 Amanita 10 4.78
小菇科 Mycenaceae 2 12 5.74 丝盖伞属 Inocybe 5 2.39
多孔菌科 Polyporaceae 8 14 6.70 蜡蘑属 Laccaria 5 2.39
红菇科 Russulaceae 3 23 11.00 小皮伞属 Marasmius 6 2.87
小菇属 Mycena 11 5.26
光柄菇属 Pluteus 5 2.39
红菇属 Russula 17 8.13
栓菌属 Trametes 5 2.39
关闭全屏