摘要
针对IRT模型的参数估计过程中传统的统计方法难以解决迭代初值要求严格等问题,提出利用遗传算法(GA)优化神经网络(BP)、共同求解IRT模型项目参数的GA-BP算法,提高收敛能力和求解精度。将GA-BP算法与BILOG算法进行对比,通过实验验证了GA-BP算法的正确性及合理性。
Focusing on the problem that Traditional statistical methods are difficult to solve the stringent requirements of initial iteration value in the parameter estimation progress of IRT models, GA - BP algorithm is proposed to estimate item parameters of IRT models. GA -BP uses genetic algorithm (GA) to optimize the neural network (BP) to improve the convergence ability and solution accuracy. The comparison result of GA - BP and BILOG shows the correctness and reasonableness of GA - BP.
出处
《华北电力大学学报(自然科学版)》
CAS
北大核心
2012年第5期109-112,共4页
Journal of North China Electric Power University:Natural Science Edition
基金
国家自然科学基金资助项目(31101078)