Journal: Scientific Reports

Article Title: A fuzzy rank-based ensemble of CNN models for classification of cervical cytology

doi: 10.1038/s41598-021-93783-8

Figure Lengend Snippet: Examples of images from the Mendeley LBC dataset . HSIL high squamous intra-epithelial lesion, LSIL low squamous intra-epithelial lesion, NIL negative for intra-epithelial lesion, SCC squamous cell carcinoma.

Article Snippet: Figure 12 Visualization of the convolution filters of the Inception v3 model on the Mendeley LBC dataset (the plots have been formed using Keras framework of Python).

Techniques:

Journal: Scientific Reports

Article Title: A fuzzy rank-based ensemble of CNN models for classification of cervical cytology

doi: 10.1038/s41598-021-93783-8

Figure Lengend Snippet: Results obtained on ensembling various combinations of base learners on all the three datasets used in this study.

Article Snippet: Figure 12 Visualization of the convolution filters of the Inception v3 model on the Mendeley LBC dataset (the plots have been formed using Keras framework of Python).

Techniques:

Journal: Scientific Reports

Article Title: A fuzzy rank-based ensemble of CNN models for classification of cervical cytology

doi: 10.1038/s41598-021-93783-8

Figure Lengend Snippet: Visualization of the convolution filters of the Inception v3 model on the Mendeley LBC dataset (the plots have been formed using Keras framework of Python).

Article Snippet: Figure 12 Visualization of the convolution filters of the Inception v3 model on the Mendeley LBC dataset (the plots have been formed using Keras framework of Python).

Techniques:

Journal: Scientific Reports

Article Title: A fuzzy rank-based ensemble of CNN models for classification of cervical cytology

doi: 10.1038/s41598-021-93783-8

Figure Lengend Snippet: Comparison of the classification performance of the base learners and their ensemble using the proposed scheme.

Article Snippet: Figure 12 Visualization of the convolution filters of the Inception v3 model on the Mendeley LBC dataset (the plots have been formed using Keras framework of Python).

Techniques:

Journal: Scientific Reports

Article Title: A fuzzy rank-based ensemble of CNN models for classification of cervical cytology

doi: 10.1038/s41598-021-93783-8

Figure Lengend Snippet: Loss curves obtained on fine-tuning the three CNN base learners: Inception v3, Xception and DenseNet-169 on the three datasets used in this research— (a–c) SIPaKMeD 2-class dataset, (d–f) SIPaKMeD 5-class dataset and (g–i) Mendeley LBC 4-class dataset (The loss curves have been plotted using Keras framework of Python).

Article Snippet: Figure 12 Visualization of the convolution filters of the Inception v3 model on the Mendeley LBC dataset (the plots have been formed using Keras framework of Python).

Techniques:

Journal: Scientific Reports

Article Title: A fuzzy rank-based ensemble of CNN models for classification of cervical cytology

doi: 10.1038/s41598-021-93783-8

Figure Lengend Snippet: Results of the McNemar’s test performed between the proposed ensemble model and the base learners used: null hypothesis is rejected for all cases.

Article Snippet: Figure 12 Visualization of the convolution filters of the Inception v3 model on the Mendeley LBC dataset (the plots have been formed using Keras framework of Python).

Techniques:

Journal: Scientific Reports

Article Title: A fuzzy rank-based ensemble of CNN models for classification of cervical cytology

doi: 10.1038/s41598-021-93783-8

Figure Lengend Snippet: The architecture of the Inception v3 model: base learner 1 (image has been made by R.K. using Google Slides).

Article Snippet: However, for the visualization purpose, we have provided the filters of convolution for the Inception v3 model on the Mendeley LBC dataset in Fig. .

Techniques:

Journal: Scientific Reports

Article Title: A fuzzy rank-based ensemble of CNN models for classification of cervical cytology

doi: 10.1038/s41598-021-93783-8

Figure Lengend Snippet: Mathematical steps of the proposed ensemble method using three CNN base models. I represents the input images; P represents the decision scores generated by the base learner and i represents the base learners: Inception v3 ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$i=1$$\end{document} i = 1 ), Xception ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$i=2$$\end{document} i = 2 ) and DenseNet-169 ( \documentclass[12pt]{minimal} \usepackage{amsmath} \usepackage{wasysym} \usepackage{amsfonts} \usepackage{amssymb} \usepackage{amsbsy} \usepackage{mathrsfs} \usepackage{upgreek} \setlength{\oddsidemargin}{-69pt} \begin{document}$$i=3$$\end{document} i = 3 ) (image has been made by R.K. using Google Slides).

Article Snippet: However, for the visualization purpose, we have provided the filters of convolution for the Inception v3 model on the Mendeley LBC dataset in Fig. .

Techniques: Generated

Journal: Scientific Reports

Article Title: A fuzzy rank-based ensemble of CNN models for classification of cervical cytology

doi: 10.1038/s41598-021-93783-8

Figure Lengend Snippet: Results obtained on ensembling various combinations of base learners on all the three datasets used in this study.

Article Snippet: However, for the visualization purpose, we have provided the filters of convolution for the Inception v3 model on the Mendeley LBC dataset in Fig. .

Techniques:

Journal: Scientific Reports

Article Title: A fuzzy rank-based ensemble of CNN models for classification of cervical cytology

doi: 10.1038/s41598-021-93783-8

Figure Lengend Snippet: Visualization of the convolution filters of the Inception v3 model on the Mendeley LBC dataset (the plots have been formed using Keras framework of Python).

Article Snippet: However, for the visualization purpose, we have provided the filters of convolution for the Inception v3 model on the Mendeley LBC dataset in Fig. .

Techniques:

Journal: Scientific Reports

Article Title: A fuzzy rank-based ensemble of CNN models for classification of cervical cytology

doi: 10.1038/s41598-021-93783-8

Figure Lengend Snippet: Comparison of the classification performance of the base learners and their ensemble using the proposed scheme.

Article Snippet: However, for the visualization purpose, we have provided the filters of convolution for the Inception v3 model on the Mendeley LBC dataset in Fig. .

Techniques: Comparison

Journal: Scientific Reports

Article Title: A fuzzy rank-based ensemble of CNN models for classification of cervical cytology

doi: 10.1038/s41598-021-93783-8

Figure Lengend Snippet: Loss curves obtained on fine-tuning the three CNN base learners: Inception v3, Xception and DenseNet-169 on the three datasets used in this research— (a–c) SIPaKMeD 2-class dataset, (d–f) SIPaKMeD 5-class dataset and (g–i) Mendeley LBC 4-class dataset (The loss curves have been plotted using Keras framework of Python).

Article Snippet: However, for the visualization purpose, we have provided the filters of convolution for the Inception v3 model on the Mendeley LBC dataset in Fig. .

Techniques:

Journal: Scientific Reports

Article Title: A fuzzy rank-based ensemble of CNN models for classification of cervical cytology

doi: 10.1038/s41598-021-93783-8

Figure Lengend Snippet: Comparison of the proposed ensemble model with some standard CNN models in literature: Inception v3 , Xception , DenseNet-169 , ResNet-18 , VGG-19 (image has been made by R.K. using Google Sheets).

Article Snippet: However, for the visualization purpose, we have provided the filters of convolution for the Inception v3 model on the Mendeley LBC dataset in Fig. .

Techniques: Comparison

Journal: Scientific Reports

Article Title: A fuzzy rank-based ensemble of CNN models for classification of cervical cytology

doi: 10.1038/s41598-021-93783-8

Figure Lengend Snippet: Comparison of the proposed ensemble model with some popular fusion techniques in literature using the same base learners: Inception v3, Xception and DenseNet-169 (image has been made by R.K. using Google Sheets).

Article Snippet: However, for the visualization purpose, we have provided the filters of convolution for the Inception v3 model on the Mendeley LBC dataset in Fig. .

Techniques: Comparison

Journal: Scientific Reports

Article Title: A fuzzy rank-based ensemble of CNN models for classification of cervical cytology

doi: 10.1038/s41598-021-93783-8

Figure Lengend Snippet: Examples of test samples from the SIPaKMeD Pap Smear dataset where one or more of the base classifiers predict incorrectly, but the ensemble predicts correctly. (a) DenseNet-169 classifies the sample as: “Koilocytotic” with confidence 31%, Xception classifies the sample as: “Parabasal” with confidence 36% and Inception v3 classifies the sample as: “Metaplastic” with confidence 98%. Ensemble prediction is: “Metaplastic”. (b) DenseNet-169 classifies the sample as: “Koilocytotic” with confidence 32%, Xception classifies the sample as “Parabasal” with confidence 95%, and Inception v3 classifies the sample as “Parabasal” with confidence 98%. Ensemble prediction is: “Parabasal”.

Article Snippet: However, for the visualization purpose, we have provided the filters of convolution for the Inception v3 model on the Mendeley LBC dataset in Fig. .

Techniques:

Journal: Scientific Reports

Article Title: A fuzzy rank-based ensemble of CNN models for classification of cervical cytology

doi: 10.1038/s41598-021-93783-8

Figure Lengend Snippet: Results of the McNemar’s test performed between the proposed ensemble model and the base learners used: null hypothesis is rejected for all cases.

Article Snippet: However, for the visualization purpose, we have provided the filters of convolution for the Inception v3 model on the Mendeley LBC dataset in Fig. .

Techniques: Comparison

Journal: Scientific Reports

Article Title: A fuzzy rank-based ensemble of CNN models for classification of cervical cytology

doi: 10.1038/s41598-021-93783-8

Figure Lengend Snippet: Results (accuracies in %) obtained by the proposed ensemble framework and its base classifiers on the Zenodo 5K breast histopathology dataset.

Article Snippet: However, for the visualization purpose, we have provided the filters of convolution for the Inception v3 model on the Mendeley LBC dataset in Fig. .

Techniques: Histopathology

Journal: Health Information Science and Systems

Article Title: Features extraction using encoded local binary pattern for detection and grading diabetic retinopathy

doi: 10.1007/s13755-022-00181-z

Figure Lengend Snippet: Performances of existing DR detection methods

Article Snippet: Sengupta et al. [ ] , EyePACS , Inception-v3 Model , Acc: 90. 4%, Sensitivity: 90%.

Techniques: Software

Journal: BMC Genomics

Article Title: Design and performance of a bovine 200 k SNP chip developed for endangered German Black Pied cattle (DSN)

doi: 10.1186/s12864-021-08237-2

Figure Lengend Snippet: Comparison of variant effects between the DSN200k SNP chip and the Illumina BovineSNP50 BeadChip using the Ensembl Variant Effect Predictor (VEP). The color indicates the impact of each consequence from the least severe (blue) to the most severe (red)

Article Snippet: The complete list of variants is provided in Table S . Altogether, SNPs of the DSN200k SNP chip had an overlap of 49,569 SNPs and 35,025 SNPs with the Illumina BovineHD BeadChip (Illumina Inc., CA, USA) and the Illumina BovineSNP50 BeadChip v.3, respectively.

Techniques: Variant Assay

Journal: BMC Genomics

Article Title: Design and performance of a bovine 200 k SNP chip developed for endangered German Black Pied cattle (DSN)

doi: 10.1186/s12864-021-08237-2

Figure Lengend Snippet: Number of unique, total, successfully called (high-quality genotype calls), and in the population segregating variants (SNPs and indels) on the DSN200k SNP chip per category of selection

Article Snippet: The complete list of variants is provided in Table S . Altogether, SNPs of the DSN200k SNP chip had an overlap of 49,569 SNPs and 35,025 SNPs with the Illumina BovineHD BeadChip (Illumina Inc., CA, USA) and the Illumina BovineSNP50 BeadChip v.3, respectively.

Techniques: Selection