Dataset with multiple features for Deep Learning toolbox

Hello,

I am trying to use the DEEPlearning toolbox , the ia importing my dataset from a CSV file which has 3 columns , the first two being the inputs and the third one is the output. I am trying to import the CSV file using the command >>tabularTextDatastore() , but when I try to train my network with the toolbox , it says that the dataset has one feature , How am I supoose to create datasets with multiple features ?

[Iuliu Ardelean]

Could you show the relevant bits of your code?

[Parham Mohammadi]

I am using the graphic interface of the toolbar and I am feeding it with data

[Parham Mohammadi]

Is there any way to define a data store for multiple scalar features? I am using the feature layer

Answer 1

[Nagasai Bharat]

Hi,

From my understanding you are trying to use the tabularTextDatastore to store the data from your CSV files. One of the reason you are not able to get the other columns is that the delimiter which seperates the entries is not specified. Another case you can chose the SelectedVariableNames Key and appropriate Values to choose the columns.

Try going through Datastore for tabular text files - MATLAB (mathworks.com) documentation.

[Parham Mohammadi]

Can you please give the exact command in order to define the dataset suitable for multiple input and multiple output neural network ?