Table 2.

Precision, Recall and F-score of our model and K-means method^∗.

No.	Our model Precision	Our model Recall	Our model F2 Score	K-means Precision	K-means Recall	K-means F2 Score
1	70.69%	98.48%	91.30%	43.08%	68.92%	61.54%
2	82.10%	99.74%	95.63%	93.61%	48.52%	53.69%
3	56.90%	91.32%	81.46%	16.20%	58.86%	38.56%
4	83.57%	99.77%	96.04%	94.00%	46.21%	51.44%
5	82.80%	99.34%	95.52%	91.17%	71.04%	74.32%
6	73.86%	98.50%	92.34%	78.70%	68.51%	70.33%
7	83.10%	99.39%	95.64%	91.50%	66.78%	70.60%
8	83.55%	99.37%	95.74%	89.44%	52.63%	57.35%
9	64.47%	97.30%	88.31%	63.78%	53.10%	54.94%
10	72.42%	98.66%	91.99%	72.42%	63.30%	64.93%
11	73.32%	99.42%	92.81%	79.89%	68.11%	70.18%
12	79.50%	99.53%	94.76%	81.04%	68.47%	70.66%
13	66.39%	95.35%	87.70%	49.58%	66.79%	62.45%
14	85.88%	99.68%	96.58%	95.43%	46.34%	51.65%
15	71.35%	89.38%	85.08%	73.48%	54.89%	57.82%
16	68.33%	97.63%	89.91%	65.84%	60.89%	61.82%
17	59.08%	86.48%	79.14%	40.65%	61.70%	55.90%
18	73.13%	99.22%	92.61%	84.52%	56.67%	60.67%
19	65.46%	98.89%	89.72%	65.23%	67.06%	66.69%
20	70.04%	99.44%	91.74%	57.37%	62.13%	61.12%

^∗The average Precision, Recall and F2 score of our model in test samples are 73.30, 97.34, and 91.20%, respectively; the average Precision, Recall and F2 score of K-means method in test samples are 71.35, 60.55, and 60.83%, respectively.