Research Article

Deep Learning Based Syndrome Diagnosis of Chronic Gastritis

Table 1

Results of model with different nodes’ number (mean ± std.).

hidAverage precision ↑Coverage Hamming loss One-error Ranking loss

50.791 ± 0.0190.191 ± 0.0130.154 ± 0.0110.304 ± 0.0290.151 ± 0.016
100.802 ± 0.0250.175 ± 0.0190.144 ± 0.0140.301 ± 0.0430.133 ± 0.021
200.802 ± 0.0290.175 ± 0.0220.145 ± 0.0120.303 ± 0.0410.133 ± 0.026
300.817 ± 0.0270.164 ± 0.0210.139 ± 0.0120.283 ± 0.0390.120 ± 0.023
400.812 ± 0.0170.168 ± 0.0160.138 ± 0.0150.292 ± 0.0290.125 ± 0.017
500.815 ± 0.0190.166 ± 0.0180.137 ± 0.0100.289 ± 0.0270.123 ± 0.018
600.816 ± 0.0230.164 ± 0.0200.135 ± 0.0090.287 ± 0.0390.121 ± 0.020
700.818 ± 0.0180.164 ± 0.0150.139 ± 0.0090.283 ± 0.0280.120 ± 0.016
800.823 ± 0.0180.158 ± 0.0150.139 ± 0.0140.278 ± 0.0280.116 ± 0.016
900.821 ± 0.0200.160 ± 0.0160.137 ± 0.0120.278 ± 0.0340.116 ± 0.017
1000.817 ± 0.0230.166 ± 0.0170.136 ± 0.0060.281 ± 0.0450.123 ± 0.018
2000.818 ± 0.0210.164 ± 0.0160.139 ± 0.0090.282 ± 0.0350.120 ± 0.018
3000.815 ± 0.0230.164 ± 0.0190.141 ± 0.0110.287 ± 0.0390.122 ± 0.022