welcom I want to load target in neural networks. How can download target in function nnstart

I want to be pinned down a goal in neural networks. How can download target in function nnstart

[Greg Heath]

Why can't you do it the same way you downloaded the input?

Greg

[fatima ali]

i use it the same way you downloaded the input but dont downloaded

[Greg Heath]

I don't understand.

[Mustafa Kadhm]

Salam Fatima,

The target represent the desired output , means something you want to represent your required classes .

Besides, the target matrix should be correspond to the input matrix through the number of columns . The number of rows will represent your classes number.

Therefore , in your situation you should create target like:

T=[ 1 1 1 0 0 0 0 0 0 0; 0 0 0 1 1 1 0 0 0 0; 0 0 0 0 0 0 0 1 1 1]; If you do not have a target like this one , nothing will appear in the target list.

I hope this helpful for you .

Regards,

[santhosh kumar buddepu]

how to create this target data? can I write manually like you mentioned? my input is 76*6 (76 images with 6 features) and I have 3 classes. means my target data should be 76*3. how to create this data.?

I have assigned labels to my database like "traindatabase.Labels" which gives 76*1 categorical data

[Bandi]

sir, Am unable to see the above window after pressing the fitting option

Answer 1

[yanqi liu]

yes，sir，may be the data should be one-hot type，such as

>> [x,t] = simpleclass_dataset;
>> nprtool

then，you will found that the t is

of course，if you want to make your own data，such as

X = rand(4, 5);
Y = zeros(3, 5);
Y(2,1:3) = 1;
Y(1,4:5) = 1;

[santhosh kumar buddepu]

thank you @yanqi liu

but if i have large dataset and many number of classes everytime shall i need to do like this. or is there any other suggestion

Y = zeros(3, 5);

Y(2,1:3) = 1;

Y(1,4:5) = 1;

[yanqi liu]

yes，sir，if the class number is 3，then use one-hot type to range target data Y