Table 7.

Performance comparison of the proposed ensemble model with the existing state-of-the-art techniques on the Databiox dataset for the classification of IDC-BC grade images.

Reference	Year	Approach	Performance Metric
Zavareh et al. [22]	2021	Transfer learning approach (VGG16 used as feature extractor)	Accuracy of 72%
Kumaraswamy et al. [25]	2021	Transfer learning approach pre-trained CNNs: DensNet201 and NASNetMobile used as feature extractors)	Accuracy of 72%. AUC for Grade 1, and Grade 2 is 98% and 75%, respectively with DensNet201 AUC for Grade 3 is 69% with NASNetMobile
Sujatha et al. [24]	2022	Transfer learning approaches (Utilized VGG16, VGG19, InceptionReNetV2, DenseNet121, and DenseNet201)	DenseNet121 produced the highest accuracy of 92.64%
Talpur et al. [23]	2022	A sequential convolutional neural network is utilised	Accuracy of 92.81%
Present Work	2023	Proposed Ensemble Model(EfficientNetV2L + ResNet152V2 + DensNet201)	Accuracy of 94%. AUC of 96%, 94% and 96% for Grades 0, 1, and 2, respectively.