# Something wrong when recognizing data set using Binary Artificial Neural Network

I have learnt Neural Network for a couple weeks. I just met a problem when I tried to use ANN to recognize a dataset.
So I'm gonna describe the problem:
Set A has 15 members which are in set {1, 1.5, 2, 2.5... 4}. I need to recognize set A if it is in type 2 or not. There are 3 types:

• Type 1: Sum(A) < 30
• Type 2: 30 <= Sum(A) <36
• Type 3: Sum(A)>=36

I'm able to recognize if A is type 1 or type 3, but something went wrong when I tried with type 2. I think because type 2 is in the middle of type 1 and type 3 so I need to do something more.

I tried 2 models of binary ANN:

1. The Neural network has 3 layers,input: 16 cells, hidden: 8 cells, output 1 cell.

2. The Neural network has 4 layers,input: 16 cells, hidden1: 8 cells, hidden2: 2 cells, output 1 cell.

Biases have been using through the network.

So the question is am I able to recognize the type 2 by using a single binary ANN? It means separating type 2 (class A) from type 1 and 3 (class B). And if it is possible, can you recommend the model of neural network?

• I guess your set is actually a list, since there is no set of size 20 whose elements come from 1, 1.5, 2, ..., 4. – David Richerby Aug 9 '16 at 9:23
• @DavidRicherby yeah, thanks. for example {1, 1.5, 2, 4, 3.5, 3.5, 1, 1, 3, 2, 1.5, 3, 3, 2, 1} – Đoàn Minh Tiến Aug 9 '16 at 10:19
• What were the results from your attempts? – William Schaller Sep 23 '16 at 15:53