Research Article

Least Absolute Deviation Support Vector Regression

Table 1

Experiment results on .

Noise Algorithm RMSE MAE SSE/SST SSR/SST

Type I LS-SVR 0.0863 0.0676 0.0720 1.0761
LAD-SVR0.04550.03640.02091.0082

Type II LS-SVR 0.0979 0.0795 0.0944 0.9374
LAD-SVR0.0727 0.0607 0.0527 1.0224

Type III LS-SVR0.0828 0.0665 0.0664 1.1137
LAD-SVR0.0278 0.0223 0.0078 1.0044

Type IV LS-SVR0.0899 0.0707 0.0772 0.9488
LAD-SVR0.0544 0.0439 0.0291 1.0003

Type V LS-SVR0.2120 0.1749 0.4268 0.7142
LAD-SVR0.1861 0.1554 0.3372 0.7291

Type VI LS-SVR0.1796 0.1498 0.3112 0.9709
LAD-SVR0.1738 0.1391 0.2928 0.8993