问题描述
我有200个样本,(在下面的链接是一个明显的草图)每个样本有4个输入(a,b,c,(t1,t2,t3))和1个输出(根据t1,t2和t3)。(t1,t2和t3)对于所有样本都不一致,即第一个样本(t1 = 3天,t2 = 7天,t3 = 15天),第二个样本(t1 = 2天,t2 = 9天) ,t3 = 17天等等.....我的问题:Narx(动态神经网络)是否有利于解决我的问题,如果不是什么是我应该使用的ANN的合适架构。我想预测输出这个范围内的输入不是将来。我注意到我使用Matlab的Toolbox而不是编程代码。
I have 200 samples,(In the bottom link is a obvious sketch)each sample has 4 Inputs (a,b,c,(t1,t2,t3))and 1 Output (variable according to the t1,t2 and t3).The(t1,t2 and t3)is not consistent for all samples i.e. for first sample(t1=3 day,t2=7 day,t3=15 day),Second sample(t1=2 day,t2=9 day,t3=17 day) and so on..... My question: Is Narx(dynamic neural network) benefit to solve my problem and if not what is a appropriate architecture of ANN that i should use.I want to prediction of output with inputs within this range not in future.I note that i use Toolbox of Matlab not a programming code.
http://www.mathworks.com/matlabcentral/answers/uploaded_files/14084/test.png
推荐答案
这篇关于在时间序列步骤不是恒定且从一个样本到另一个样本不一致的情况下使用NARX的文章就介绍到这了,希望我们推荐的答案对大家有所帮助,也希望大家多多支持!