Title :
Marginal maximum likelihood estimation of single parameter logistic based on EM algorithm
Author :
Sun, Xueyan ; Jing, Fengxuan ; Xie, Xiaoyao ; Zhang, Anyu
Author_Institution :
Key Lab. of Inf. & Comput. Sci. of Guizhou Province, Guizhou Normal Univ., Guiyang, China
Abstract :
Cluster analysis is one of the most important functions of data mining. Expectation Maximization (EM) method is an important technology based on model clustering method. The expectation maximization algorithm is analyzed in this research and applied to Adaptive Testing System, in which logistic function in item response theory serves as a model, and the combination of methods of marginal maximum likelihood estimation (MMLE) and the EM algorithm are used to estimate the difficulty parameter estimation of single-parameter logistic function. This effort achieves good results.
Keywords :
data analysis; data mining; expectation-maximisation algorithm; pattern clustering; adaptive testing system; cluster analysis; data mining; expectation maximization algorithm; item response theory; logistic function; marginal maximum likelihood estimation; model clustering method; parameter estimation; single parameter logistic; Adaptation model; Clustering algorithms; Data models; Logistics; Maximum likelihood estimation; Presses; EM algorithm; logistic function; parameter estimation;
Conference_Titel :
Anti-Counterfeiting Security and Identification in Communication (ASID), 2010 International Conference on
Conference_Location :
Chengdu
Print_ISBN :
978-1-4244-6731-0
DOI :
10.1109/ICASID.2010.5551506
