Skip to main content
. 2021 Oct 20;4:762899. doi: 10.3389/fdata.2021.762899

TABLE 2.

Overall Performance with respect to RMSE, MAE, MAPE and MSLE. (The smaller value is, the better performance is).

RMSE Outperform MAE Outperform MAPE Outperform MSLE Outperform
AutoFTP 18.646 16.192 58.851 0.2267
AttentionWalk 21.418 +14.9% 19.712 +21.7% 68.590 +16.6% 0.2907 +28.2%
ProNE 21.830 +17.1% 19.929 +23.1% 69.188 +17.6% 0.2949 +30.1%
GatNE 21.229 +13.9% 19.288 +19.1% 67.043 +13.9% 0.2854 +25.9%
GAE 21.338 +14.4% 19.676 +21.5% 68.579 +16.5% 0.2894 +27.6%
DeepWalk 23.561 +26.4% 21.987 +35.8% 76.038 +29.2% 0.3321 +46.5%
Node2Vec 22.688 +21.7% 21.084 +30.2% 73.135 +24.3% 0.3152 +39.0%
Struc2Vec 21.589 +15.8% 19.937 +23.1% 69.423 +17.9% 0.2942 +29.7%
AutoFTP R 21.965 +17.8% 20.283 +25.3% 70.991 +20.6% 0.2928 +29.1%
AutoFTP(R+P) 20.509 +9.99% 18.921 +16.8% 66.477 +12.9% 0.2681 +18.3%
AutoFTP(R+C) 21.014 +12.7% 19.413 +19.8% 67.920 +15.4% 0.2773 +22.3%
AutoFTP(R+P+C) 20.211 +8.39% 18.676 +15.3% 65.685 +11.6% 0.2636 +16.3%