基于网络分布式随机多任务优化算法
Random Multi-Task Optimization Algorithm Based on Distributed Network
摘要: 在大数据环境下，需要对海量数据进行分析处理，加快数据处理效率。分布式网络在处理进程任务时，通过调配网络中节点资源，分配给不同的节点处理不同计算、通信任务。无中心分布式一致优化问题大部分都是以无约束为基础，即表示每个节点的初始化节点为空。本文应用梯度投影法分布式方法，提出了随机优化的策略，融合本地目标的优化操作和邻居值，将融合获取的结果都投影给本地约束集。实验结果表明本算法稠密网络中信息融合的速度更快。
Abstract: Under the big data working environment, it is necessary to analyze and process massive data to speed up the efficiency of data processing. The distributed network handles the tasks of the process by allocating the resources of the nodes in the network and assigning them to different nodes to handle different computing and communication tasks. Most of the decentralized and consistent optimization problems are based on unconstrained, which means that the initialization node of each node is empty. In this paper, a gradient projection method distributed method is used, and a stochastic optimization strategy is proposed. The optimization operation of the local target and the neighbor values are merged, and the results obtained by the fusion are projected to the local constraint set. Experimental results show that the proposed algorithm is faster in information fusion in dense networks.
文章引用：陈创明, 温洁嫦. 基于网络分布式随机多任务优化算法[J]. 应用数学进展, 2020, 9(6): 838-843. https://doi.org/10.12677/AAM.2020.96100

参考文献

[1] Koren, Y., Bell, R. and Volinsky, C. (2009) Matrix Factorization Techniques for Recommender systems. IEEE Computer, 42, 30-37. [Google Scholar] [CrossRef] 
[2] You, Q., Wu, O., Luo, G., et al. (2016) A Probabilistic Matrix Factorization Method for Link Sign Prediction in Social Networks. Machine Learning and Data Mining in Pattern Recognition, 9729, 15-420. [Google Scholar] [CrossRef] 
[3] Shahriari, M., Sichani, O.A., Gharibshah, J., et al. (2016) Sign Prediction in Social Networks Based on Users Reputation and Optimism. Social Network Analysis and Mining, 6, 91. [Google Scholar] [CrossRef] 
[4] Javari, A., Qiu, H.X., Barzegaran, E., et al. (2017) Statistical Link Label Modeling for Sign Prediction: Smoothing Sparsity by Joining Local and Global Information. Proceedings of the 17th IEEE International Conference on Data Mining, New Orleans, LA, 18-21 November 2017, 1039-1044. [Google Scholar] [CrossRef] 
[5] Stanimirovi, P.S. and Miladinovi, M.B. (2010) Accelerated Gradient Descent Methods with Line Search. Numerical Algorithms, 54, 503-520.
[6] Gemulla, R., Nijkamp, E., Haas, P.J., et al. (2011) Large-Scale Matrix Factorization with Distributed to Chastic Gradient Descent. Proceedings of the 17th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining, August 2011, 69-77. [Google Scholar] [CrossRef] 
[7] Chin, W.S., Zhuang, Y., Juan, Y.C., et al. (2015) A Fast Parallel Stochastic Gradient Method for Matrix Factorization in Shared Memory Systems. ACM Transactions on Intelligent Systems and Technology, 6, 2. [Google Scholar] [CrossRef] 
[8] Zhang, H., Hsieh, C.J. and Akella, V. (2016) Hogwild++: A New Mechanism for Decentralized Asynchronouss to Chastic Gradient Descent. Proceedings of the 16th IEEE International Conference on Data Mining, Barcelona, 12-15 December 2016, 629-638.57. [Google Scholar] [CrossRef] 

为你推荐