Train recurrent neural network with variable length input cell vectors.

10 次查看（过去 30 天）
Essa Anas
Essa Anas 2016-11-30
回答： Essa Anas 2016-12-1
Hi,
I am working on recurrent neural network and my dataset consists of input cell and target cell. The input cell consists of vectors with variable length for example input{[1,1,0],[2,1,1,1,2,1],[2,1],...} and my target is T{[0,0,1],[0,0,0,0,0,1],[0,1],....} I tried to use the command [Xs,Xi,Ai,Ts] = preparets(net,input,T); but Matlab through the error (Error using preparets (line 78)
Inputs{1,3} and Inputs{1,1} have different numbers of columns)
Which command should I use to overcome this issue?
Thanks in advance for any advice.

请先登录，再回答此问题。

采纳的回答

Greg Heath
Greg Heath 2016-12-1
Neural networks require "I"nput column vectors with a fixed length (I). Therefore you will have to complete the columns with zeros.
[ I N ] = size(input)
[ O N ] = size(target)
Hope this helps.
Thank you for formally accepting my answer
Greg

更多回答（1 个）

Essa Anas
Essa Anas 2016-12-1
Thank you for your reply. Yes, after changing the input to be with equal length, I was able to train the RNN.

类别

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

标签

产品

Community Treasure Hunt

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

Start Hunting!

Translated by