应用非线性加权的集成学习软件缺陷序列预测算法

贾晓琳; 樊帅帅; 罗雪; 朱晓燕

doi:10.7652/xjtuxb201707022

您当前的位置：

首页 >

文章列表页 >

应用非线性加权的集成学习软件缺陷序列预测算法

更新时间：2025-07-09

- 应用非线性加权的集成学习软件缺陷序列预测算法
- Prediction Algorithm for Software Defect Series Based on Nonlinear Weighted Ensemble Learning
- 西安交通大学学报 2017年51卷第7期页码：156-161
- 作者机构：
  
  西安交通大学电子与信息工程学院学院,西安,710049
- 作者简介：
- 基金信息：
  
  : 2017-01-12。作者简介: 贾晓琳(1963—),女,高工,硕士生导师。基金项目: 国家自然科学基金资助项目(61402355);中央高校基本科研业务费专项基金资助项目(jj2014050)
- DOI：10.7652/xjtuxb201707022
  中图分类号： O121.8
- 网络首发：2017-07-10，
  
  纸质出版：2017
- 稿件说明：
移动端阅览
贾晓琳, 樊帅帅, 罗雪, 等. 应用非线性加权的集成学习软件缺陷序列预测算法[J]. 西安交通大学学报, 2017,51(7):156-161.

Prediction Algorithm for Software Defect Series Based on Nonlinear Weighted Ensemble Learning[J]. 2017, 51(7): 156-161.
贾晓琳, 樊帅帅, 罗雪, 等. 应用非线性加权的集成学习软件缺陷序列预测算法[J]. 西安交通大学学报, 2017,51(7):156-161. DOI： 10.7652/xjtuxb201707022.

Prediction Algorithm for Software Defect Series Based on Nonlinear Weighted Ensemble Learning[J]. 2017, 51(7): 156-161. DOI： 10.7652/xjtuxb201707022.

摘要

针对当前软件缺陷序列预测算法准确度不高的问题

提出了基于非线性加权的集成学习软件缺陷序列预测算法(NLWEPrediction)。该算法在常见线性集成预测算法的基础上增加了非线性回归项

回归项代表了集成预测算法中基预测算法之间的相互关系

修正了线性集成预测的偏差

并通过梯度下降法求解了模型中的参数。实验表明:NLWEPrediction在14个软件缺陷数据集上的均方误差均小于250

并且平均绝对误差均小于13。通过与基预测算法、集成预测Bagging、Stacking算法和只考虑两个分类器关系的非线性加权集成学习算法进行对比

可以看出

NLWEPrediction预测算法的均方误差和平均绝对误差显著减小

预测精度显著提高

说明在线性集成预测算法基础上增加非线性回归项

能够有效提高集成学习算法的分类效果。

Abstract

Aiming at the problem of the basic prediction algorithm with relative low prediction accuracy

a novel and improved prediction algorithm NLWEPrediction is proposed based on non-linear weighted and ensemble learning. It combines the advantages of linear ensemble learning and the relationship between the base predict algorithms

which corrects the prediction deviation and uses gradient descent method to calculate the model parameters. The experiments proved that the NLWEPrediction's mean squared error in datasets is lower than 250

and the mean absolute difference is lower than 13. The algorithm was compared with its four base prediction algorithms

other two ensemble prediction algorithms Bagging and Stacking and original NLWEPrediction for efficiency analysis. Experimental results showed that NLWEPrediction has obviously low mean square error and average absolute error. The prediction accuracy is improved. So

adding the nonlinear regression terms can improve the capability of ensemble classifier.

关键词

Keywords

references

HE Liang, SONG Qinbao, SHEN Junyi. Boosting-based k-NN learning for software defect prediction [J]. Pattern Recognition and Artificial Intelligence, 2012, 25(5): 792-802.

RZTEBLATT Ä D. Linear regression analysis: part 14 of a series on evaluation of scientific publications [J]. Deutsches Ärzteblatt Archiv Medizin Übersichtsarbeit, 2010, 107(44): 776-782.

UMAR S N. Software testing defect prediction model: a practical approach [J]. International Journal of Research in Engineering and Technology, 2013, 2(5): 741-745.

YANG J F, ZHAI Y J, XU D P, et al. SMO algorithm applied in time series model building and forecast [C]∥International Conference on Machine Learning and Cybernetics. Piscataway, NJ, USA: IEEE, 2007: 2395-2400.

ROBERTS S, OSBORNE M, EBDEN M, et al. Gaussian processes for time-series modelling [J]. Philosophical Transactions of the Royal Society A, 2013, 371(1984): 20110550.

AUDIBERT J Y. Bagging predictors [J]. Annales De Linstitut Henri Poincare Probability Statistics, 2004, 40(6): 685-736.

SILL J, TAKACS G, MACKEY L, et al. Feature-weighted linear stacking [EB/OL]. [2016-12-12]. https:∥arxiv.org/pdf/0911.0460.pdf.

ADHIKARI R, AGRAWAL R K. A novel weighted ensemble technique for time series forecasting [J]. Lecture Notes in Computer Science, 2012, 7301: 38-49.

NESTEROV Y. Introductory lectures on convex optimization [M]. Berlin, Germany: Springer, 2014: 16-38.

浏览量

下载量

CSCD

文章被引用时，请邮件提醒。

提交

工具集

关联资源

图像分割的谱聚类集成算法

一种新的选择性支持向量机集成学习算法

一种自相似业务量预测的卡尔曼滤波算法

用于木马流量检测的集成分类模型

不平衡工艺参数数据集的高温透平叶片铸件质量预测方法