Identification of Functionally Interconnected Neurons Using Factor Analysis

Table 3

(a) Average loading matrices obtained for the interconnection scheme of Figure 1(d) using the model of correlated current. These were calculated for different weighting values. Two hundred simulations were realized for each situation. , , and fire independently (in italic font) and neuron turns out to be presynaptic to the others (in bold font). (b) Idem (a) using neural network model. (c) Theoretical loading matrix for the interconnection scheme of Figure 1(d).

(a)

0.4		0.5		0.6		0.7		0.8
0.0991	0.1425	0.0891	0.1396	0.1199	0.1348	0.1187	0.1162	0.0906	0.1252
0.0908	0.1416	0.1202	0.1888	0.0411	0.1881	0.0849	0.1668	0.0252	0.1818
0.1409	0.1768	0.1194	0.1703	0.0384	0.1497	0.0736	0.2176	0.0652	0.1699
0.4898	0.2621	0.6305	0.2257	0.6563	0.2866	0.6827	0.2726	0.7660	0.2773
0.4836	0.3162	0.6131	0.2428	0.6634	0.2752	0.6945	0.2844	0.7548	0.2783

(b)

0.4		0.5		0.6		0.7		0.8
0.0177	0.1103	0.0002	0.1450	0.0128	0.1025	0.0021	0.1213	0.0080	0.1557
0.0211	0.0886	−0.0065	0.1089	−0.0205	0.1207	0.0013	0.0988	−0.0386	0.1895
0.0021	0.1352	−0.0087	0.1229	−0.0046	0.1646	−0.0298	0.1399	0.0139	0.0893
0.7149	0.1680	0.7720	0.1635	0.7584	0.1827	0.7799	0.1899	0.8223	0.1570
0.6886	0.1702	0.7427	0.1590	0.8068	0.1620	0.8133	0.2037	0.8303	0.1491

(c)

0	0
0	0
0	0
1	0
1	0