Research Article

Robust Extreme Learning Machine Using New Activation and Loss Functions Based on M-Estimation for Regression and Classification

Table 4

Testing RMSE and SD of existing and proposed weighted ELM in clean abalone data.

AfELMW1ELMW2ELMW3ELM
Proposed
W4ELM
Proposed
W5ELM
Proposed
W6ELM Proposed

Sigmoid2.144143 0.14429432.175000 0.1369262.348306 0.309631772.205588 0.160945352.191642 0.135862632.173392 0.13340782.189162 0.138817
Tan-sig2.150399 0.19267982.123554 0.16439632.298845 0.244190122.157511 0.201615242.135203 0.161915672.126287 0.16260332.132039 0.164313
Sine2.186667 0.20097612.129157 0.14520622.330158 0.151040442.161264 0.159712862.139394 0.135026652.133264 0.14507952.136425 0.137740
Cosine2.213637 0.20631202.186199 0.17764032.384910 0.195149642.248835 0.201665152.196252 0.166587652.185097 0.17193002.195070 0.1720428
BentIdle2.132008 0.16056042.129026 0.15610232.277284 0.097419512.156375 0.175020702.137727 0.149204652.132926 0.15596772.134048 0.1507907
RAF2.145623 0.14633262.161926 0.14265742.374974 0.17473441.235957 0.234725742.180706 0.149630182.159569 0.13377372.178774 0.1537632
Proposed 12.166058 0.18741002.131701 0.148462.284179 0.244889472.157106
0.15988542
2.137113 0.138832882.136875 0.14880992.134278 0.1413268
Proposed 22.151613 0.16614442.133518 0.15767402.250627 0.263010002.151150 0.181359242.138221 0.149180272.138806 0.15752922.135509 0.1522103
Proposed 32.380380 0.31869602.286334 0.23519612.428068 0.289078432.324140 0.269643482.283338 0.219562502.282506 0.22787872.286572 0.2275768
Proposed 42.192119 0.20366082.161120 0.17758932.224148 0.126171832.154010 0.175953572.159759 0.165759192.168203 0.18078832.157113 0.1676623
Proposed 52.145925 0.15350602.162594 0.14688362.397043 0.308588322.238244 0.232967312.183591 0.155989562.162214 0.14040162.180518 0.1585908
Proposed 62.139137 0.16341562.162571 0.18879722.399246 0.239874762.212400 0.215153372.179009 0.186788142.160730 0.17995382.176938 0.1914957
Proposed 72.321377 0.27320222.195224 0.18839432.358959 0.114845802.245618 0.256991352.210419 0.190941632.193319 0.17685872.207753 0.1976999
Proposed 82.167783 0.20024442.633435 0.10511622.609206 0.117526552.850021 0.028059032.748345 0.076663152.616541 0.10891422.730395 0.0814038
Proposed 92.155725 0.17380432.143697 0.13831952.321712 0.168595672.195728 0.191722202.158484 0.139407782.145574 0.13382232.155555 0.1426305
Proposed 102.130854 0.15755252.134321 0.15071162.179927 0.122464462.158516 0.130179822.145262 0.149177002.136971 0.15354752.142324 0.1481731
Proposed 112.138887 0.14721352.120890 0.13898612.242633 0.053011442.164488 0.182330372.129773 0.130059452.125429 0.13967722.126699 0.1324753