how to train data

I'm starting to learn neural network and I'd like to know if it's possible to do the following:

I have text data, like this:

1 apple, 1.23, 5.26, fruit

2 carrot, 4.56, 7.25, vegetable

3 banana, 1.34, 6.77, fruit

4 orange, 1.96, 6.88, fruit

5 potato, 7.23, 2.33, vegetable

and so on...

apple, carrot... are the names of the images... next two column values are features extracted... and the last column is the text i wanted to get displayed when the particular image is selected....

just the values in the last column are different, but all fitting a simple pattern.

I'd like to use this as training data, and then input to the network a question like:

if apple is selected fruit should be displayed?

Can that be done in matlab? Please could someone reply? I can't even enter the data properly... if without using neural networks also please do someone reply....

Answer 1

[Walter Roberson]

Why bother with a network? Why not just have a look-up table,

T = {'apple', 'fruit'; 'carrot', 'vegetable'; 'banana', 'fruit'}
name = tolower( input('Enter name of item', 's') );
[tf, idx] = ismember(name, T(:,1));
if tf
  fprintf('%s is a kind of %s\n', name, T{idx, 2});
else
  fprintf('I do not know what a "%s" is\n', name);
end

[Elysi Cochin]

is it possible using a mat file?... if yes, how to do using a mat file??

[Walter Roberson]

Read the data in. save() the useful parts of it as a cell array in a .mat file. load() the .mat file when you need it.