[Retracted] Rapid Identification of Tobacco Mildew Based on Random Forest Algorithm

Table 9

Training results of random forest model based on different wavelet coefficients.

Wavelet coefficient	Variable length	Recognition rate/%	Recognition rate 1/%	Recognition rate 2/%	Recognition rate 3/%
cd1	1051	78.53	77.81	45.59	85.88
cd2	540	80.85	81.10	58.82	85.00
cd3	284	82.79	80.82	58.82	89.71
cd4	156	90.94	90.68	80.83	93.24
cd5	92	84.22	90.14	73.53	80.00
cd6	60	82.28	87.95	72.06	78.24
cd7	44	81.76	83.29	70.59	82.35
cd8	36	79.69	83.56	70.59	77.35
cd9	32	79.04	79.45	58.82	82.65
cd10	30	74.77	81.37	36.76	75.29
cd11	29	73.48	79.73	35.29	74.41
ca11	29	59.51	67.67	27.94	57.06
[cd4, cd5]	248	93.40	93.42	91.17	93.82
Primitive harmonic	2074	69.73	73.15	30.88	73.82