Skip to main content

Table 6 Performance evaluation based on six different models

From: Examining the impact of cross-domain learning on crime prediction

Season Model Random forest Gradient boosting
AUC (%) Gmean (%) AUC (%) Gmean (%)
1 1 69.69 69.25 69.25 68.95
  2 64.34 63.39 65.17 65.14
  3 63.18 62.61 63.17 62.69
  4 69.55 69.13 70.25 69.73
  5 70.00 69.48 70.05 69.74
  6 69.69 69.11 70.10 69.71
2 1 69.61 69.57 69.25 69.24
  2 64.47 64.35 65.04 64.94
  3 63.09 62.10 63.74 63.54
  4 69.54 69.50 70.63 70.59
  5 69.41 69.35 70.35 70.35
  6 69.28 69.20 70.52 70.52
3 1 67.57 67.55 67.54 67.49
  2 63.74 63.65 63.86 63.66
  3 62.89 62.72 62.54 62.06
  4 67.42 67.41 68.70 68.68
  5 67.57 67.56 67.99 67.87
  6 67.52 67.51 68.12 68.04
4 1 68.76 68.68 68.12 68.08
  2 64.20 64.08 64.80 64.54
  3 62.85 62.84 62.41 62.37
  4 68.65 68.54 69.23 69.09
  5 68.79 68.66 68.99 68.96
  6 68.80 68.66 69.11 69.07