Overestimation Trap of Artificial Neural Network: Learning the Rule of PRBS
Liang Shu, Jianqiang Li, Zhiquan Wan, Wenjia Zhang, Songnian Fu, Kun Xu · 2018
We investigate the overestimation trap of ANN from learning the rule of PRBS over unrepeated pattern. The influence of input symbols and hidden neurons is studied by simulations. Besides, the overestimated gain is experimentally evaluated.