Research Article
[Retracted] Design of Optimal Deep Learning-Based Pancreatic Tumor and Nontumor Classification Model Using Computed Tomography Scans
Table 1
Comparative classification results analysis of ODL-PTNTC technique under varying training set (TS).
| Training (%) | ODL-PTNTC | DS-WELM | DS-KELM | DS-ELM |
| Sensitivity | TS = 40 | 99.89 | 99.69 | 96.97 | 96.79 | TS = 50 | 98.55 | 98.35 | 98.23 | 97.49 | TS = 60 | 98.32 | 97.20 | 98.00 | 97.12 | TS = 70 | 97.59 | 96.53 | 97.02 | 97.13 | TS = 80 | 99.31 | 98.89 | 97.82 | 96.28 | Average | 98.73 | 98.13 | 97.61 | 96.96 |
| Specificity | TS = 40 | 96.96 | 96.22 | 96.87 | 96.96 | TS = 50 | 97.20 | 96.84 | 97.12 | 95.43 | TS = 60 | 98.82 | 98.70 | 95.78 | 97.78 | TS = 70 | 97.57 | 96.46 | 97.19 | 95.31 | TS = 80 | 98.18 | 97.53 | 97.64 | 97.74 | Average | 97.75 | 97.15 | 96.92 | 96.64 |
| Accuracy | TS = 40 | 98.34 | 98.29 | 95.44 | 97.46 | TS = 50 | 99.08 | 97.60 | 97.92 | 98.93 | TS = 60 | 98.50 | 98.47 | 94.87 | 94.59 | TS = 70 | 97.21 | 96.52 | 97.02 | 94.32 | TS = 80 | 98.86 | 97.42 | 98.37 | 96.90 | Average | 98.40 | 97.66 | 96.72 | 96.44 |
| F-score | TS = 40 | 98.92 | 98.39 | 98.59 | 94.32 | TS = 50 | 99.08 | 95.76 | 98.84 | 96.51 | TS = 60 | 99.84 | 99.70 | 97.10 | 97.08 | TS = 70 | 98.27 | 96.71 | 95.55 | 97.88 | TS = 80 | 97.98 | 96.40 | 94.31 | 97.93 | Average | 98.82 | 97.39 | 96.88 | 96.74 |
|
|