摘要
提出一种基于CNN算法的网络信任等级模型,解决用户节点信任值的高低对整个网络信任度造成影响问题。使用设定好的卷积神经网络对用户节点数据进行训练,完成用户信任等级分类工作,在提高执行效率的基础上,对数据完成有效分类,确定用户特征属性的具体维度范围,并通过噪声、运行时间、准确率和分类精度等不同方面对不同算法进行对比验证模型,实验结果表明,改进后的算法在集群环境下执行的效率得到很大提升,能够高效处理实验数据。
In paper,a convolutional neural network(CNN) based network trust model is proposed,aiming to solve the problem of how trust level of users affect the trust level of the whole network. We use configured convolutional neural network to train the user node data to complete the classification of user trust level,on the basis of improving the efficiency of the implementation,the data is classified effectively and the specific dimension range of user characteristics is determined. To verify the model,different algorithms are compared by different aspects such as noise,running time,precision and classification accuracy. The experimental results show that the efficiency of the improved algorithm in the cluster environment is greatly improved,and the experimental data can be processed efficiently.
出处
《长春理工大学学报(自然科学版)》
2017年第4期93-98,共6页
Journal of Changchun University of Science and Technology(Natural Science Edition)
基金
吉林省科技发展计划重点科技攻关项目(20150204036GX)