Performance Analyses of Discrete-Time RNN for Solving Discrete-Form Time-Variant Matrix Inversion with Different Selection Parameters
2022 IEEE 11th Data Driven Control and Learning Systems Conference (DDCLS)(2022)
摘要
Neural network could be considered as a basic artificial intelligence methods. In this paper, we explore a lot of researches on performance analyses of discrete-time recurrent neural network (DT-RNN) model. For solving the discrete-form time-variant matrix inversion (DF-TV-MI), continuous-time recurrent neural network (CT-RNN) model is presented firstly. Then, an inspirational method named general-four-instant discretization formula (GFI discretization formula) to discrete the CT-RNN model, and we obtain a new DT-RNN model. Finally, we show the performance analyses of DT-RNN model solving for DF-TV-MI with different selection parameters.
更多查看译文
关键词
different selection parameters,matrix,discrete-time,discrete-form,time-variant
AI 理解论文
溯源树
样例
生成溯源树,研究论文发展脉络
Chat Paper
正在生成论文摘要