Fig 1.
Data modalities in healthcare.
Fig 2.
Traditional data fusion.
Fig 3.
The proposed FKG-MM framework.
Fig 4.
FKG-MM for diagnosis of diabetic retinopathy diseases.
Fig 5.
FKG of 6 fuzzy rules [65].
Table 1.
An illustration of 6 fuzzy rules [23].
Table 2.
FRB consists of 14 fuzzy rules [65].
Table 3.
FKGS consists of 5 fuzzy rules [65].
Fig 6.
Heatmap tabular feature.
Fig 7.
Heatmap image feature.
Fig 8.
Feature importance chart.
Table 4.
Performance evaluation metrics.
Table 5.
Feature selection method with sampling ratio of 15% and error threshold of 0.2.
Table 6.
Feature selection method with sampling ratio of 15% and error threshold of 0.3.
Table 7.
Feature selection method with sampling ratio 20% and error threshold 0.2.
Table 8.
Feature selection method with sampling ratio 20% and error threshold 0.3.
Fig 9.
Comparison of accuracy.
Table 9.
ANOVA results for accuracy.
Table 10.
FKG-MM results with sampling ratio of 15% and error threshold of 0.2.
Fig 10.
Confusion matrix of Feature Selection method with sample rate 20% and error threshold 0.3.
Table 11.
FKG-MM results with sampling ratio of 15% and error threshold of 0.3.
Table 12.
FKG-MM results with sampling ratio of 20% and error threshold of 0.2.
Table 13.
FKG-MM results with sampling ratio of 20% and error threshold of 0.3.
Table 14.
ANOVA results for accuracy.
Table 15.
ANOVA results for total time.
Table 16.
Tukey HSD results for accuracy.
Table 17.
Tukey HSD results for total time.