[1]康春花,张淑君,李元白,等.KNN认知诊断法及其应用[J].江西师范大学学报(自然科学版),2019,(02):135-141+159.[doi:10.16357/j.cnki.issn1000-5862.2019.02.04]
　KANG Chunhua,ZHANG Shujun,LI Yuanbai,et al.The Cognitive Diagnosis of k-Nearest Neighbor Algorithm and Its Application[J].Journal of Jiangxi Normal University:Natural Science Edition,2019,(02):135-141+159.[doi:10.16357/j.cnki.issn1000-5862.2019.02.04]
点击复制

KNN认知诊断法及其应用()
分享到：

《江西师范大学学报》（自然科学版）[ISSN:1006-6977/CN:61-1281/TN]

卷:
期数:
2019年02期
页码:
135-141+159
栏目:
心理与教育测量
出版日期:
2019-04-10

文章信息/Info

Title:
The Cognitive Diagnosis of k-Nearest Neighbor Algorithm and Its Application
文章编号:
1000-5862(2019)02-0135-07
作者:
康春花; 张淑君; 李元白; 曾平飞*
浙江师范大学教师教育学院,浙江 金华 321004
Author(s):
KANG Chunhua; ZHANG Shujun; LI Yuanbai; ZEBNG Pingfei*
College of Teacher Education,Zhejiang Normal University,Jinhua Zhejiang 321004,China
关键词:
KNN算法;  KNN认知诊断法;  实证信效度
Keywords:
KNN algorithm;  KNN CDM;  empirical reliability and validity
分类号:
B 841
DOI:
10.16357/j.cnki.issn1000-5862.2019.02.04
文献标志码:
A
摘要:
将机器学习中的KNN算法迁移至认知诊断评估中,提出了KNN认知诊断法,并通过模拟和实证研究考察了KNN认知诊断法的效果和特征.结果表明:KNN认知诊断法具有较高的判准率,与PNN和MDD-R诊断法不相上下,甚至在某些情境下更高; KNN认知诊断法不受样本容量和被试知识状态分布形态的影响,体现了KNN算法作为非参数方法的特征; KNN认知诊断法具有较好的实证信效度.
Abstract:
In the study,the k-Nearest Neighbors cognitive diagnosis method(KNN CDM)is proposed by migrating the k-Nearest Neighbors algorithm to cognitive diagnosis assessment.Then its effectiveness and characteristic are investigated by mumerical simulation and empirical study.The results show that the precision of the KNN CDM is high,and is about the same to PNN method as well as MDD-R method.In some situations,it's even higher than the latter two.Both the sample size and the distribution pattern of the subjects' knowledge state have no effect on the KNN CDM,which reflects the characteristic of nonparametric method.It is proved that the emprical reliability and validity of the KNN cognitive diagnosis method is good.

参考文献/References:

[1] 袁梅宇.数据挖掘与机器学习WEKA应用技术与实践[M].北京:清华大学出版社,2014.
[2] 何清,李宁,罗文娟,等.大数据下的机器学习算法综述[J].模式识别与人工智能,2014,27(4):327-336.
[3] 邝铮.支持向量机在认知诊断中的应用研究[D].南昌:江西师范大学,2010.
[4] Liu Cheng,Cheng Ying.An application of the support vector machine for attribute-by-attribute classification in cognitive diagnosis[J].Applied Psychological Measurement,2018,42(1):58-72.
[5] 汪文义,丁树良,宋丽红,等.神经网络和支持向量机在认知诊断中的应用[J].心理科学,2016,39(4):777-782.
[6] 唐小娟.粗糙集理论在认知诊断中的应用[D].南昌:江西师范大学,2013.
[7] Harrington P.Machine learning in action[J].Shelter Island,NY:Manning Publications Co,2012.
[8] Tan Pangning,Michael Steinbach,Vipin Kumar.Introduction to data mining[M].New Jersey:Addison Wesley,2005.
[9] 桑应宾.基于k近邻的分类算法研究[D].重庆:重庆大学,2009.
[10] Cover T,Hart P.Nearest neighbor pattern classification[J].IEEE Transactions on Information Theory,1967,13(1):21-27.
[11] Chiu Chiayi,Köhn H F.A general proof of consistency of heuristic classification for cognitive diagnosis models[J].British Journal of Mathematical and Statistical Psychology,2015,68(3):387-409.
[12] Chiu Chiayi,Köhn H F.Consistency of cluster analysis for cognitive diagnosis:the reduced reparameterized unified model and the general diagnostic model[J].Psychometrika,2016,81(3):585-610.
[13] Köhn H F,Chiu Chiayi,Brusco M J.Heuristic cognitive diagnosis when the Q-matrix is unknown[J].British Journal of Mathematical and Statistical Psychology,2015,68(2):268-291.
[14] 罗欢,丁树良,汪文义,等.属性不等权重的多级评分属性层级方法[J].心理学报,2010,42(4):528-538.
[15] 田伟,辛涛.基于等级反应模型的规则空间方法[J].心理学报,2012,44(2):249-262.
[16] 任平.多级聚类诊断法的开发及其在认知诊断评估中的应用[D].金华:浙江师范大学,2015.
[17] 康春花,任平.聚类诊断分析法诊断正确率的影响因素[J].中国考试,2015(2):25-32.
[18] 康春花,杨亚坤,曾平飞.海明距离判别法分类准确率的影响因素[J].江西师范大学学报:自然科学版,2017,41(4):394-400.
[19] 杨亚坤.观察反应模式与理想反应模式间曼哈顿距离判别法[D].金华:浙江师范大学,2017.
[20] Chiu Chiayi,Douglas J A,Li Xiaodong.Cluster analysis for cognitive diagnosis:theory and applications[J].Psychometrika,2009,74(4):633.
[21] Chiu Chiayi,Douglas J.A nonparametric approach to cognitive diagnosis by proximity to ideal response patterns[J].Journal of Classification,2013,30(2):225-250.
[22] 罗照盛,李喻骏,喻晓锋,等.一种基于Q矩阵理论朴素的认知诊断方法[J].心理学报,2015,47(2):264-272.
[23] 康春花,辛涛,田伟.小学数学应用题认知诊断测验编制及效度验证[J].考试研究,2013(6):24-43.
[24] Roussos L A,DiBello L V,Stout W,et al.The fusion model skills diagnosis system[EB/OL].[2018-06-19].http://pdfs.semanticscholar.org/c757/afd523c9f31063e4c
efea84176762763ee73.pdf.
[25] 戴晓阳,曹亦薇.心理评定量表的编制和修订中存在的一些问题[J].中国临床心理学杂志,2009,17(5):562-565.
[26] 蔡艳,涂冬波,丁树良.五大认知诊断模型的诊断正确率比较及其影响因素:基于分布形态、属性数及样本容量的比较[J].心理学报,2013,45(11):1295-1304.
[27] 丁树良,罗芬,汪文义.Q矩阵理论的扩展[J].心理学探新,2012,32(5):417-422.
[28] 颜远海,丁树良,汪文义.影响AHM与DINA诊断准确率的因素研究[J].江西师范大学学报:自然科学版,2011,35(6):640-645.
[29] DiBello L V,Stout W.Guest editors' introduction and overview:IRT-based cognitive diagnostic models and related methods[J].Journal of Educational Measurement,2007,44(4):285-291.
[30] Leighton J P,Gierl M J,Hunka S M.The attribute hierarchy method for cognitive assessment:a variation on Tatsuoka's rule-space approach[J].Journal of Educational Measurement,2004,41(3):205-237.

备注/Memo

备注/Memo:
收稿日期:2018-10-17 基金项目:教育部人文社会科学研究一般课题(16YJA190002)资助项目. 通信作者:曾平飞(1963-),男,广西荔浦人,教授,博士,主要从事心理测量与评价方面的研究.E-mail:zpf@zjnu.edu.cn
更新日期/Last Update: 2019-04-10