Title :
Improved LMBP algorithm in the analysis and application of simulation data
Author :
Xue, Qing ; Yun, Fei ; Zheng, Changwei ; Liu, Yonghong ; Wei, Yang ; Yao, Yijun ; Zhou, Shaoliang
Author_Institution :
Acad. of Armored Force Eng., Beijing, China
Abstract :
Based on the analysis of features and reasoning course of LMBP(Levenberg-Marquardt) algorithm, we propose the way of variable learning rate to improve the convergence speed. Using some literature data of reference standard database and program MATLAB simulation experiment, it shows that this improved the feasibility and effectiveness LMBP algorithm. Improving LMBP algorithm offers important reference value for combat simulation equipment for the analysis of the data characteristics and summary conclusions.
Keywords :
convergence; data analysis; learning (artificial intelligence); mathematics computing; military computing; Levenberg-Marquardt algorithm; combat simulation equipment; convergence speed; improved LMBP algorithm; program MATLAB simulation; reference standard database; simulation data analysis; variable learning rate; Algorithm design and analysis; BP network; LMBP algorithm; MATLAB; combat simulation;
Conference_Titel :
Computer Application and System Modeling (ICCASM), 2010 International Conference on
Print_ISBN :
978-1-4244-7235-2
Electronic_ISBN :
978-1-4244-7237-6
DOI :
10.1109/ICCASM.2010.5620688
