1. 1. 毕业设计(论文)的内容、要求、设计方案、规划等
Plane-Gaussian人工神经网络,由于采用Plane-Gaussian函数作为隐层结点激活函数,使该网络模型同时具备全局性和局部性能,而且从理论上也证明了该类型网络亦具有全局逼近能力。与BP、RBF神经网络相比,实验中验证了该网络具有相当的分类能力。然而,因为PG网络先需要通过k平面聚类,并根据聚类结果决定隐层结点数和计算输出权矩阵,如此不此训练速度较慢,而且容易陷入局部极小点。针对此类问题,本文拟采用随机投影的方法,这样不仅能够保证PG函数中的超平面方向选择更具多样性,有利于突破局部极小点的限制,而且神经网络训练速度应该更快。在阅读相关文献的基础上,实现上述功能,并按照南京林业大学毕业论文撰写规范,完成一篇本科毕业论文。
2. 参考文献(不低于12篇)
[1] X. Yang, S. Chen, et.al. Plane-Gaussian artificial neural network. Neural computing and applications, 2012, 21(2): 305-317.[2] X. Z. Fern, C.E. Brodley. Random projection for high dimensional data clustering: a cluster ensemble approach. Proceedings of the Twentieth International Conference on
Machine Learning (ICML), 2003[3] Bradley PS, Mangasarian OL. k-plane clustering. J Global Optimization, 2000,16(1):2332[4] Guang-Bin Huang, Qin-Yu Zhu, Chee-Kheong Siew. Extreme Learning machine: Theory and applications. Neurocomputing, 2006, 70:489-501.[5] T. Watanabe, E. Takimoto, et.al. Random projection and its application to learning. Joint workshop "new horizons in computing" and statistical mechanical approach to
probabilistic information processing", 2005[6] X Liu, S.Lin, J.Fang and Zongben Xu. Is extreme learning machine Feasible? a theoretical assessment. IEEE trans. on Neural networks and learning systems, 2015, 26(1):7-34[7] R.Livni, S.S Shai, O.Shamir. On the computational efficiency of training neural networks. NIPS, 2014.[8] G.Huang, G.Huang, S Song, K. You. Trends in extreme learning machines: A review. Neural Networks, 2015, 61:32-48[9] 王颖, 陈松灿,张道强等. 模糊k-平面聚类算法.模式识别与人工智能, 2007, 20(5): 704-710 [10] 邓万宇, 郑庆华,陈琳,许学斌. 神经网络极速学习方法研究. 计算机学报。2010, 33(2): 279-287
以上是毕业论文任务书,课题毕业论文、开题报告、外文翻译、程序设计、图纸设计等资料可联系客服协助查找。