There are many kinds of big data in peer to peer communication network, traditional big data integration algorithms ignore the classification of big data state. The location correlation between big data cannot be accurately obtained, which leads to the time-consuming process of big data integration and the low acquisition accuracy of big data location correlation. A fuzzy integration algorithm for big data in peer-to-peer communication network based on deep learning is proposed. According to the big data integration conditions of peer-to-peer communication network, the benchmark model of big data integration is constructed by combining semi-supervised deep learning. The marked sample updating model of the prior single classification fuzzy integrated data model is used to process the corresponding sample data. Combining the steady-state and dynamic residuals of the datum model, multiple information grains are obtained. Based on this, kalman filter is adopted to fuzzy fusion of big data to obtain state parameters, and multi-scale analysis is carried out to filter out big data noise. Using the obtained information grains to constrain the characteristic correlation degree of big data, the position correlation of big data is completed. Implement fuzzy integration algorithm design of big data in peer-to-peer communication network based on deep learning. In order to verify the effectiveness of the proposed algorithm, a simulation experiment is designed. Experimental results show that compared with many traditional methods, the proposed algorithm takes less time and the computational complexity of the proposed algorithm is significantly reduced. And the big data integration is more accurate and has better application effect.

对等通信网络中的大数据种类繁多，传统的大数据集成算法忽略了大数据状态的分类。大数据之间的位置相关性无法准确获得，导致大数据集成的过程耗时，大数据位置相关性的获取精度也较低。提出了一种基于深度学习的对等通信网络中大数据的模糊集成算法。根据对等通信网络的大数据集成条件，结合半监督深度学习，构建大数据集成的基准模型。先前的单分类模糊综合数据模型的标记样本更新模型用于处理相应的样本数据。结合基准模型的稳态和动态残差，获得了多个信息晶粒。在此基础上，采用卡尔曼滤波器对大数据进行模糊融合以获得状态参数，并进行多尺度分析以滤除大数据噪声。利用获得的信息粒度约束大数据的特征相关度，完成了大数据的位置相关。基于深度学习的对等通信网络中实现大数据模糊集成算法设计。为了验证所提算法的有效性，设计了一个仿真实验。实验结果表明，与许多传统方法相比，该算法耗时少，计算复杂度大大降低。