基于广义化自适应的NAG方法非负张量分解模型
Non-Negative Tensor Decomposition ModelBased on Generalized and Adaptive NAG Method
摘要: 单元素非负乘法更新算法在学习模型超参数时会出现长尾收敛的情况，本文通过将NAG方法融入到单元素非负乘法更新算法中，得到了广义化的NAG方法，并在此基础上提出了基于广义化自适应的NAG非负张量分解模型。在训练过程中利用粒子群算法对模型的正则化系数和算法的加速度系数进行了优化。最后，在两个真实的工业数据集上的对比实验表明，本文提出的广义化NAG方法明显提高了模型的收敛速度。
Abstract: In order to solve the problem of long tail convergence when SLF-NMU algorithm learns model hy-perparameters, the NAG method is integrated into the SLF-NMU algorithm, and the generalized NAG method is obtained. On this basis, a non-negative tensor decomposition based on generalized and adaptive NAG method is proposed. During the training process, particle swarm optimization is used to optimize parameters. Comparison with three similar algorithms on real industrial data shows that the proposed model can improve the convergence speed.
文章引用：陶名康. 基于广义化自适应的NAG方法非负张量分解模型[J]. 应用数学进展, 2022, 11(5): 3134-3149. https://doi.org/10.12677/AAM.2022.115333

参考文献

[1] Ghahramani, M.H., Zhou, M.C. and Hon, C.T. (2017) Toward Cloud Computing QoS Architecture: Analysis of Cloud Systems and Cloud Services. IEEE/CAA Journal of Automatica Sinica, 4, 6-18. [Google Scholar] [CrossRef] 
[2] Fareghzadeh, N. (2022) An Architecture Supervisor Scheme to-ward Performance Differentiation and Optimization in Cloud Systems. The Journal of Supercomputing, 78, 1532-1563. [Google Scholar] [CrossRef] 
[3] Luo, X., Zhou, M.C., Xia, Y., et al. (2015) Generating Highly Accurate Predictions for Missing QoS Data via Aggregating Nonnegative Latent Factor Models. IEEE Transactions on Neural Networks and Learning Systems, 27, 524-537. [Google Scholar] [CrossRef] 
[4] Luo, X., Zhou, M.C., Li, S., et al. (2017) Incorporation of Ef-ficient Second-Order Solvers into Latent Factor Models for Accurate Prediction of Missing QoS Data. IEEE Transac-tions on Cybernetics, 48, 1216-1228. [Google Scholar] [CrossRef] 
[5] Chang, Z., Ding, D. and Xia, Y. (2021) A Graph-Based QoS Prediction Approach for Web Service Recommendation. Applied Intelligence, 51, 6728-6742. [Google Scholar] [CrossRef] 
[6] Koren, Y. (2010) Collaborative Filtering with Temporal Dy-namics. Communications of the ACM, 53, 89-97. [Google Scholar] [CrossRef] 
[7] Luo, X., Wu, H., Yuan, H., et al. (2019) Temporal Pattern-Aware QoS Prediction via Biased Non-Negative Latent Factorization of Tensors. IEEE Transactions on Cybernetics, 50, 1798-1809. [Google Scholar] [CrossRef] 
[8] Nesterov, Y.E. (1983) A Method of Solving a Convex Pro-gramming Problem with Convergence Rate O (1/k2). Soviet Mathematics Doklady, 27, 372-376.
[9] Attouch, H., Bolte, J. and Svaiter, B.F. (2013) Convergence of Descent Methods for Semi-Algebraic and Tame Problems: Proximal Algo-rithms, Forward-Backward Splitting, and Regularized Gauss-Seidel Methods. Mathematical Programming, 137, 91-129. [Google Scholar] [CrossRef] 
[10] Luo, X., Liu, Z., Li, S., et al. (2018) A Fast Non-Negative Latent Factor Model Based on Generalized Momentum Method. IEEE Transactions on Systems, Man, and Cybernetics: Sys-tems, 51, 610-620. [Google Scholar] [CrossRef] 
[11] Sutskever, I., Martens, J., Dahl, G., et al. (2013) On the Im-portance of Initialization and Momentum in Deep Learning. International Conference on Machine Learning, Atlanta, 16-21 June 2013, 1139-1147.
[12] Aggarwal, C.C. (2016) Recommender Systems. Springer International Publishing, Cham. [Google Scholar] [CrossRef] 
[13] Javed, K., Gouriveau, R. and Zerhouni, N. (2015) A New Multi-variate Approach for Prognostics Based on Extreme Learning Machine and Fuzzy Clustering. IEEE Transactions on Cy-bernetics, 45, 2626-2639. [Google Scholar] [CrossRef] 
[14] Shi, Y. (2001) Particle Swarm Optimization: Developments, Applications and Resources. Proceedings of the 2001 Congress on Evolutionary Computation, Seoul, 27-30 May 2001, 81-86.
[15] Wijnhoven, R.G.J. and de With, P.H.N. (2010) Fast Training of Object Detection Using Stochastic Gradient Descent. 2010 20th International Conference on Pattern Recognition, Istanbul, 23-26 August 2010, 424-427. [Google Scholar] [CrossRef] 

为你推荐