I am taking images of numbers as input, in a convolutional neural network and building a model to predict the number.

In particular, I am building a one dimensional convolutional neural network with single channel to take images of numbers (one dimensional arrays) as input. This would be Case 1.

In another case, I take the same input arrays and I split each input by half and stack the two halves on top to get a two dimensional array representation of each input. I build a one dimensional convolutional neural, which has two channels for two dimensional array inputs. I call this Case 2.

How would the accuracies of predictions of case 1 and case 2 compare with each other?

Please not that I am referring to all cases that I am referring to here are 1dcnn with 1 channel (Case 1) or two channels (Case 2). I am not referring to 2dcnn in any of these cases.



You must log in to answer this question.