Single Neuron for Solving XOR like Nonlinear Problems

Table 7

Scaling Factor and loss Obtained by Π_t-Neuron and Proposed Models with Increasing Input Dimension of N-bit Parity Problem.

N	Scaling factor		L1 loss
	πt-neuron (bπ‒t)	Ours (b)	πt-neuron	Ours
2	‒1.7045	4.6900	0.3281	0.0093
5	0.3911	4.2760	2.1059	0.0093
7	72.1092	4.2208	0.8367	0.0058
10	‒42.6576	4.0720	0.5690	0.0047
13	33.9146	4.0548	1.0349	0.0037
15	64.4870	3.9516	0.8966	0.0079
18	‒20.1572	3.8317	0.9393	0.0072
20	‒79.3860	3.9426	0.8311	0.0072
25	83.9456	3.7275	0.9855	0.1431