| 基于多项式回归的自构形算法 |
| |
| 引用本文： | 饶泓,扶名福,陈炼,马永生. 基于多项式回归的自构形算法[J]. 南昌大学学报(理科版), 2007, 31(2): 192-196 |
| |
| 作者姓名： | 饶泓 扶名福 陈炼 马永生 |
| |
| 作者单位： | 南昌大学,计算中心,江西,南昌,330031;南昌大学,机电工程学院,江西,南昌,330031 |
| |
| 摘 要： | 自构形算法是BP神经网络提高泛化能力的一种方法,但自构形算法以线性回归理论为基础,在解决非线性问题时剪枝能力和收敛速度都存在不足,针对这一问题,提出基于多项式回归的自构形算法,并以风机故障诊断数据为例,对两种算法进行了详细的实验对比,实验表明基于多项式回归的自构形算法剪枝能力和收敛速度均优于线性自构形算法.
|
| 关 键 词： | 多项式回归 自构形 泛化 神经网络 |
| 文章编号： | 1006-0464（2007）02-0192-05 |
| 修稿时间： | 2006-12-20 |
The Self-configuring Algorithm Based on Polynomial Regression |
| |
| RAO Hong,FU Ming-fu,CHEN Lian,MA Yong-sheng. The Self-configuring Algorithm Based on Polynomial Regression[J]. Journal of Nanchang University(Natural Science), 2007, 31(2): 192-196 |
| |
| Authors: | RAO Hong FU Ming-fu CHEN Lian MA Yong-sheng |
| |
| Affiliation: | a. Center of Computer; b. School of Mechanical and Electrical, Nanchang University, Nanchang 330031, China |
| |
| Abstract: | In view of the insufficiency of the linear self - configuration algorithm in solving the non - linear problem, a self- configuring algorithm based on polynomial regression is proposed. It takes a factory air blower fault diagnosis data as the example to train the network in two algorithms. The experiment shows that both the prune ability and the convergence rate of the new algorithm surpass the linear self- configuration algorithm. |
| |
| Keywords: | polynomial regress self - configuring algorithm generalization neural network |
| 本文献已被 CNKI 维普 万方数据 等数据库收录！ |
|
