Table 9 Improved models performance for pipeline residual strength prediction
From: Machine learning methods for predicting residual strength in corroded oil and gas steel pipes
Source | Model | R2 | MSE | RMSE | MAE | MAPE |
|---|---|---|---|---|---|---|
FFNN | - | 4.650 | - | - | - | |
PSO-FFNN | - | 1.850 | - | - | - | |
SVM | - | - | 5.189 | - | 23.776% | |
PSO-SVM | - | - | 5.385 | - | 25.738% | |
WOA-SVM | - | - | 5.779 | - | 31.520% | |
NSGA-II–SVM | - | - | 0.769 | - | 4.142% | |
SVM | 0.735 | - | 4.231 | 1.726 | 14.987% | |
PSO-SVM | 0.986 | - | 1.984 | 1.437 | 9.772% | |
MOGWO-SVM | 0.999 | - | 0.315 | 0.237 | 1.353% | |
NSGA-II-SVM | 0.997 | - | 0.760 | 0.437 | 3.220% | |
ELM | 0.287 | - | 4.704 | 3.172 | 13.570% | |
TLBO-ELM | 0.692 | - | 3.571 | 3.084 | 15.912% | |
HTLBO-ELM | 0.885 | - | 2.434 | 1.856 | 8.362% | |
DELM | 0.312 | - | 4.256 | 3.188 | 13.707% | |
TLBO-DELM | 0.968 | - | 1.923 | 1.436 | 6.726% | |
HTLBO-DELM | 0.992 | - | 0.525 | 0.418 | 2.244% | |
ELM | 0.97556 | 0.70227 | - | - | - | |
GA-ELM | 0.99648 | 0.10598 | - | - | - | |
BPNN | - | 1.2830 | - | - | 8.510% | |
PSO-BPNN | - | 0.8051 | - | - | 4.840% | |
IPSO-BPNN | - | 0.6721 | - | - | 3.760% |
