So I am new to NN and I'm trying to go deep and apply it to my subject. I would like to ask: the input of the NN can be 2 or more values for example-> the measurement of a value, distance, and time? An example of input data would be [ [1,2,3, ....],[11,22,33, .....],[5] ] whose output is a value 1 for example or cat or an generated model.
$\begingroup$
$\endgroup$
1
-
$\begingroup$ Hello and welcome to this community :) To clarify what your question is, Can you please put it in the title? "About the Input in a NN" is very vague and not a question. $\endgroup$– nbroCommented Mar 16, 2021 at 11:26
Add a comment
|
1 Answer
$\begingroup$
$\endgroup$
11
Yes, it can be those multiple inputs, you would have to design the network like that (to have 3 input layers and then pool them afterwards).
But, an easier solution would be to flatten this array and put it through 1D CNN and finally put it through a softmaxx.
-
$\begingroup$ in order to have 3 input layers, i have to set the input_shape as (,3), yes? $\endgroup$– TheGameCommented Mar 17, 2021 at 7:52
-
$\begingroup$ No you need to define 3 Input layers. $\endgroup$ Commented Mar 17, 2021 at 11:51
-
$\begingroup$ pls, edit your answer with an example for 3 input layers $\endgroup$– TheGameCommented Mar 17, 2021 at 12:17
-
$\begingroup$ It is better to stack them vertically and then apply 1D CNN. I don't recommend the multiple input format; it is not prevalent and haven't seen it anywhere. But, it is possible. $\endgroup$ Commented Mar 18, 2021 at 18:43
-
$\begingroup$ but they are different data, they dont represent the same measure $\endgroup$– TheGameCommented Mar 18, 2021 at 18:48