How to add additional information in a neural network other than putting as input data?

1 次查看（过去 30 天）
Ayesha Munira Chowdhury
编辑： Ayesha Munira Chowdhury 2017-5-18
Hi! I am trying to use neural network to predict final concrete strength. I have used material properties of concrete as input data. Now if I want to add some additional information like early age strength, is there any way to put them in the hidden layer? I mean other than using as input data?
Sorry if this sounds silly, but I only have little knowledge about neural networks. 
Also, which kind of neural network works best for prediction or estimation?
Thanks in advance :) 

请先登录，再回答此问题。

采纳的回答

Greg Heath
Greg Heath 2017-5-17
No.
All design (training and validation) data info must be implemented in the input/target information before training.
Although testing can be included during training, it will have no affect on the design.
Hope this helps.
Thank you for formally accepting my answer
Greg
  1 个评论
Ayesha Munira Chowdhury
Ayesha Munira Chowdhury 2017-5-18
编辑：Ayesha Munira Chowdhury 2017-5-18
Thanks very much for your answer! Can you please suggest me which kind of neural network works best for prediction or estimation? I have used feed forward back propagation neural network before, but somebody suggested me to use NARX neural network. But from the sample example that is available in Matlab, it looks like it needs time series data rather than using some other information. Can you please clarify that also?

请先登录，再进行评论。

更多回答（0 个）

类别

在 Help Center 和 File Exchange 中查找有关 Deep Learning Toolbox 的更多信息

标签

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by