CNN Yöntem Seçimi | TOPSİS PROJE
powered by Emirhan PERVANLAR
|
Training Epoch (dk) | Resim Başına Sınıflandırma (sn) | Parametre (m) | Hata (%) | RAM (gb) | GPU (%) | Kullanım |
| GoogleNet | 4 | 2.8 | 4 | 6.66 | 4.4 | 84 | 9 |
|---|
| AlexNet | 2 | 1 | 61 | 16.4 | 2.2 | 88 | 6 |
|---|
| LeNet | 0.4 | 0.5 | 100 | 9 | 4.8 | 82 | 2 |
|---|
| VGG | 2.4 | 2 | 138 | 7.5 | 6.6 | 92 | 1 |
|---|
| ResNet | 3.2 | 1.2 | 1.7 | 6.43 | 3.9 | 76 | 2 |
|---|
| AĞIRLIK |
0.21 | 0.21 | 0.06 | 0.3 | 0.03 | 0.03 | 0.16 |
Normalize Karar Matrisi
|
Training Epoch (dk) | Resim Başına Sınıflandırma (sn) | Parametre (m) | Hata (%) | RAM (gb) | GPU (%) | Kullanım |
| GoogleNet |
0.66519010523774 |
0.73455653058677 |
0.022091731152472 |
0.30028330193964 |
0.42734562210023 |
0.44417718044971 |
0.80178372573727 |
|---|
| AlexNet |
0.33259505261887 |
0.2623416180667 |
0.33689890007519 |
0.73943635913065 |
0.21367281105012 |
0.46532847475684 |
0.53452248382485 |
|---|
| LeNet |
0.066519010523774 |
0.13117080903335 |
0.55229327881179 |
0.40578824586438 |
0.46619522410934 |
0.43360153329615 |
0.17817416127495 |
|---|
| VGG |
0.39911406314264 |
0.5246832361334 |
0.76216472476028 |
0.33815687155365 |
0.64101843315035 |
0.48647976906397 |
0.089087080637475 |
|---|
| ResNet |
0.53215208419019 |
0.31480994168004 |
0.0093889857398005 |
0.28991315787867 |
0.37878361958884 |
0.40187459183545 |
0.17817416127495 |
|---|
Ağırlıklı Normalize Karar Matrisi
|
Training Epoch (dk) | Resim Başına Sınıflandırma (sn) | Parametre (m) | Hata (%) | RAM (gb) | GPU (%) | Kullanım |
| GoogleNet |
0.13968992209993 |
0.15425687142322 |
0.0013255038691483 |
0.090084990581893 |
0.012820368663007 |
0.013325315413491 |
0.12828539611796 |
|---|
| AlexNet |
0.069844961049963 |
0.055091739794007 |
0.020213934004512 |
0.2218309077392 |
0.0064101843315035 |
0.013959854242705 |
0.085523597411976 |
|---|
| LeNet |
0.013968992209993 |
0.027545869897004 |
0.033137596728708 |
0.12173647375932 |
0.01398585672328 |
0.013008045998884 |
0.028507865803992 |
|---|
| VGG |
0.083813953259955 |
0.11018347958801 |
0.045729883485617 |
0.1014470614661 |
0.01923055299451 |
0.014594393071919 |
0.014253932901996 |
|---|
| ResNet |
0.11175193767994 |
0.066110087752809 |
0.00056333914438803 |
0.0869739473636 |
0.011363508587665 |
0.012056237755064 |
0.028507865803992 |
|---|
Max ve Min Matrisi
| S-Max |
0.13968992209993 | 0.15425687142322 | 0.045729883485617 | 0.2218309077392 | 0.01923055299451 | 0.014594393071919 | 0.12828539611796 |
| S-Min |
0.013968992209993 | 0.027545869897004 | 0.00056333914438803 | 0.0869739473636 | 0.0064101843315035 | 0.012056237755064 | 0.014253932901996 |
İdeal ve Negatif İdeal Çözümler
| S1+ | 0.13918130850804 | S1- | 0.2119376446731 |
|---|
| S2+ | 0.13174384810603 | S2- | 0.16594105374315 |
|---|
| S3+ | 0.22808873363508 | S3- | 0.050308914248645 |
|---|
| S4+ | 0.18044398800832 | S4- | 0.11885951776839 |
|---|
| S5+ | 0.19697891031791 | S5- | 0.1061904578951 |
|---|
Çözüm Matrisi
| Alternatif |
C(i) |
Tercih |
| GoogleNet | 0.60360639251441 | 1 |
|---|
| AlexNet | 0.55743859601997 | 2 |
|---|
| LeNet | 0.18070883368115 | 5 |
|---|
| VGG | 0.39712036603095 | 3 |
|---|
| ResNet | 0.35026776788507 | 4 |
|---|