Applying Randomness Effectively Based on Random Forests for Classification Task of Datasets of Insufficient Information

Table 9

The accuracy of random forests for the data set “Dermatology.”

Number of attr. to pick randomly (R)	Worst accuracy (%)	Median accuracy (%)	Best accuracy (%)	Number of trees in random forests resulting in the best accuracy (t)
12	95.9016	96.4481	96.4481	All except 100, 125, 500, 725, 800, 850 ~ 1000
6 (default)	95.6284	96.1749	96.4481	950
3	96.4481	96.9945	97.2678	200, 225, 250
2	96.7213	96.9945	97.2678	425
1	96.4481	97.2678	97.541	500, 525, 575, 600, 625,; 750 ~ 950