Page 84 - Big Data Analytics for Intelligent Healthcare Management

4.6 RESULT AND ANALYSIS 77

Xception+KNN
Xception+SVC

Xception+LR
IRV2+KNN
IRV2+SVC

IRV2+LR Recall
Precision
IV3+KNN
Accuracy
IV3+SVC
IV3+LR

RN50+KNN
RN50+SVM

RN50+LR
82 84 86 88 90 92 94 96 98 100

FIG. 4.20
Test performance graph for 100 .

Table 4.15 Result of ResNet-50 with LR, SVM, and K-NN on 200×

Feature Extractor Classifier Accuracy (%) Precision (%) Recall (%) F1-Score (%)
ResNet-50 LR 94.29 94.89 96.65 95.76
SVM 91.81 90.41 98.14 94.12
K-NN 92.06 91.01 97.31 94.05

4.6.3.5 Test performance on 200×
Interpretation: With ResNet50, SVM had the highest recall value but LR had the highest accuracy,
precision, and f1score (Table 4.15, Fig. 4.21).
Interpretation: With InceptionV3, Support Vector had the highest accuracy, recall, and f1score but
K-NN had the highest precision (Table 4.16, Fig. 4.22).
Interpretation: With Inception ResNet V2, the Support Vector classifier had the highest accuracy,
precision, recall, and f1score (Table 4.17, Fig. 4.23).
Interpretation: With Xception, the Support Vector classifier had the best recall value, K-NN had the
best precision, and LR had the highest accuracy and f1score (Table 4.18, Fig. 4.24).
Overall performance on 200