DocumentCode :
584474
Title :
The Application of GA-BP Algorithm in 3PML Parameter Estimation
Author :
Hua, Wang ; Jing, Chen ; Cuiqin, Ma
Author_Institution :
Inf. Eng. Inst., Capital Normal Univ., Beijing, China
fYear :
2012
fDate :
11-13 Aug. 2012
Firstpage :
1537
Lastpage :
1540
Abstract :
For IRT Model´s 3PLM parameter estimation, traditional statistical methods cannot solve the problems in the iterative process properly. In this paper, both BILOG algorithm and GA_BP algorithm was applied to a series of project parameters calculations to compare the two methods. The result shows that the application of GA_BP algorithm significantly reduced the initial iteration value required in traditional parameters estimation methods and markedly improved the precision and convergence ability of project parameters. Our investigation suggests GA_BP algorithm is a better method in parameter values estimation with wide applications and great research value.
Keywords :
backpropagation; convergence; education; genetic algorithms; iterative methods; parameter estimation; psychology; statistical analysis; 3 parameters logistic model; 3PML parameter estimation; BILOG algorithm; GA-BP algorithm; convergence ability; initial iteration value; item response theory; iterative process; parameter values estimation; project parameters calculations; statistical methods; Algorithm design and analysis; Educational institutions; Encoding; Estimation; Genetic algorithms; Neural networks; Parameter estimation; BILOG algorithm; GA_BP algorithm; Item Response Theory; Parameters estimation;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Computer Science & Service System (CSSS), 2012 International Conference on
Conference_Location :
Nanjing
Print_ISBN :
978-1-4673-0721-5
Type :
conf
DOI :
10.1109/CSSS.2012.385
Filename :
6394624
Link To Document :
بازگشت