Deep-layer model processing method and device
作者： Guo Zhimao , Jin Xing , Xue Wei , Li Yi , Zou Yongqiang
DOI:
关键词:
摘要: The invention discloses a deep-layer model processing method and device. comprises the following steps of: obtaining superficial-layer which is constructed trained for target event; constructing an input layer of by utilizing characteristics weighted value model; further determining number hidden layers nodes in each layer; initializing values model, so that results output are same with or similar to model. According method, initialized, those priori knowledge leaned, based on training data, used reference, quality initialized preliminary
lens.org 参考文章(4)
Yun Tang, Xing Fan, Venkatesh Nagesha, Method and Apparatus for Speech Recognition Using Neural Networks with Speaker Adaptation ,(2013)
Wu Zhuoran, Chen Yongkang, Zhang Jian, Hou Zhenjie, Chen Chen, Lu Zhongqiu, Three-dimensional face recognition algorithm based on deep learning SDAE theory and application thereof in field of finance ,(2015)
Jiang Lei, Li Yong, Zhang Shubin, Cai Bin, Yan Yanhua, Xiao Lei, Hu Shaofeng, Xue Wei, Liu Jinjing, Huang Pipei, Guo Weizhao, Liu Dapeng, Xu Miao, Wang Yang, Click stream data online analyzing method and related device and system ,(2014)
Hu Qinghua, Zhang Rujia, Short-term wind speed forecasting method based on deep neural network transfer model ,(2015)