Table 7 AUC and Gmean scores based on Toronto and Halifax data (model 4)
From: Examining the impact of cross-domain learning on crime prediction
Model 4 | Gradient boosting | TrAdaBoost | TrResampling | |||
|---|---|---|---|---|---|---|
Season | AUC (%) | Gmean (%) | AUC (%) | Gmean (%) | AUC (%) | Gmean (%) |
1 | 70.25 | 69.73 | 65.46 | 65.40 | 66.70 | 64.88 |
2 | 70.63 | 70.59 | 65.46 | 65.16 | 66.50 | 65.14 |
3 | 68.70 | 68.68 | 64.17 | 63.80 | 64.20 | 63.40 |
4 | 69.23 | 69.09 | 65.02 | 64.94 | 65.30 | 64.58 |