Explainable diabetes classification using hybrid Bayesian-optimized TabNet architecture

Article


Joseph, Lionel P., Joseph, Erica A. and Prasad, Ramendra. 2022. "Explainable diabetes classification using hybrid Bayesian-optimized TabNet architecture." Computers in Biology and Medicine. 151 (Part A). https://doi.org/10.1016/j.compbiomed.2022.106178
Article Title

Explainable diabetes classification using hybrid Bayesian-optimized TabNet architecture

ERA Journal ID5040
Article CategoryArticle
AuthorsJoseph, Lionel P., Joseph, Erica A. and Prasad, Ramendra
Journal TitleComputers in Biology and Medicine
Journal Citation151 (Part A)
Article Number106178
Number of Pages22
Year2022
PublisherElsevier
Place of PublicationUnited Kingdom
ISSN0010-4825
1879-0534
Digital Object Identifier (DOI)https://doi.org/10.1016/j.compbiomed.2022.106178
Web Address (URL)https://www.sciencedirect.com/science/article/abs/pii/S0010482522008861
Abstract

Diabetes is a deadly chronic disease that occurs when the pancreas is not able to produce ample insulin or when the body cannot use insulin effectively. If undetected, it may lead to a host of health complications. Hence, accurate and explainable early-stage detection of diabetes is essential for the proper administration of treatment options in leading a healthy and productive life. For this, we developed an interpretable TabNet model tuned via Bayesian optimization (BO). To achieve model-specific interpretability, the attention mechanism of TabNet architecture was used, which offered the local and global model explanations on the influence of the attributes on the outcomes. The model was further explained locally and globally using more robust model-agnostic LIME and SHAP eXplainable Artificial Intelligence (XAI) tools. The proposed model outperformed all benchmarked models by obtaining high accuracy of 92.2% and 99.4% using the Pima Indians diabetes dataset (PIDD) and the early-stage diabetes risk prediction dataset (ESDRPD), respectively. Based on the XAI results, it was clear that the most influential attribute for diabetes classification using PIDD and ESDRPD were Insulin and Polyuria, respectively. The feature importance values registered for insulin was 0.301 (PIDD) and for polyuria 0.206 was registered (ESDRPD). The high accuracy and ancillary interpretability of our objective model is expected to increase end-users trust and confidence in early-stage detection of diabetes.

KeywordsAttention mechanism ; “Black-box” models ; Diabetes classification ; Bayesian optimization ; eXplainable artificial intelligence (XAI) ; TabNet; Interpretability
ANZSRC Field of Research 20204699. Other information and computing sciences
Public Notes

Files associated with this item cannot be displayed due to copyright restrictions.

Byline AffiliationsSchool of Mathematics, Physics and Computing
University of Fiji, Fiji
Permalink -

https://research.usq.edu.au/item/z022x/explainable-diabetes-classification-using-hybrid-bayesian-optimized-tabnet-architecture

  • 14
    total views
  • 0
    total downloads
  • 2
    views this month
  • 0
    downloads this month

Export as

Related outputs

Short-term wind speed forecasting using an optimized three-phase convolutional neural network fused with bidirectional long short-term memory network model
Joseph, Lionel P., Deo, Ravinesh C., Casillas-Perez, David, Prasad, Ramendra, Raj, Nawin and Salcedo-sanz, Sancho. 2024. "Short-term wind speed forecasting using an optimized three-phase convolutional neural network fused with bidirectional long short-term memory network model." Applied Energy. 359. https://doi.org/10.1016/j.apenergy.2024.122624
Near real-time wind speed forecast model with bidirectional LSTM networks
Joseph, Lionel P., Deo, Ravinesh C., Prasad, Ramendra, Salcedo-Sanz, Sancho, Raj, Nawin and Soar, Jeffrey. 2023. "Near real-time wind speed forecast model with bidirectional LSTM networks." Renewable Energy. 204, pp. 39-58. https://doi.org/10.1016/j.renene.2022.12.123
Short-term wave power forecasting with hybrid multivariate variational mode decomposition model integrated with cascaded feedforward neural networks
Ali, Mumtaz, Prasad, Ramendra, Jamei, Mehdi, Malik, Anurag, Xiang, Yong, Abdulla, Shahab, Deo, Ravinesh C., Farooque, Aitazaz A. and Labban, Abdulhaleem H.. 2023. "Short-term wave power forecasting with hybrid multivariate variational mode decomposition model integrated with cascaded feedforward neural networks." Renewable Energy. 221. https://doi.org/https://doi.org/10.1016/j.renene.2023.119773
New achievements on daily reference evapotranspiration forecasting: Potential assessment of multivariate signal decomposition schemes
Ali, Mumtaz, Jamei, Mehdi, Prasad, Ramendra, Karbasi, Masoud, Xiang, Yong, Cai, Borui, Abdulla, Shahab, Farooque, Aitazaz Ahsan and Labban, Abdulhaleem H.. 2023. "New achievements on daily reference evapotranspiration forecasting: Potential assessment of multivariate signal decomposition schemes." Ecological Indicators. 155. https://doi.org/10.1016/j.ecolind.2023.111030
Ensemble robust local mean decomposition integrated with random forest for short-term significant wave height forecasting
Ali, Mumtaz, Prasad, Ramendra, Xiang, Yong, Jamei, Mehdi and Yaseen, Zaher Mundher. 2023. "Ensemble robust local mean decomposition integrated with random forest for short-term significant wave height forecasting." Renewable Energy. 205, pp. 731-746. https://doi.org/https://doi.org/10.1016/j.renene.2023.01.108
Multi-step daily forecasting of reference evapotranspiration for different climates of India: A modern multivariate complementary technique reinforced with ridge regression feature selection
Malik, Anurag, Jamei, Mehdi, Ali, Mumtaz, Prasad, Ramendra, Karbasi, Masoud and Yaseen, Zaher Mundher. 2022. "Multi-step daily forecasting of reference evapotranspiration for different climates of India: A modern multivariate complementary technique reinforced with ridge regression feature selection." Agricultural Water Management. 272. https://doi.org/https://doi.org/10.1016/j.agwat.2022.107812
Variational mode decomposition based random forest model for solar radiation forecasting: New emerging machine learning technology
Ali, Mumtaz, Prasad, Ramendra, Xiang, Yong, Khan, Mohsin, Farooque, Aitazaz Ahsan, Zong, Tianrui and Yaseen, Zaher Mundher. 2021. "Variational mode decomposition based random forest model for solar radiation forecasting: New emerging machine learning technology." Energy Reports. 7, pp. 6700-6717. https://doi.org/10.1016/j.egyr.2021.09.113
Daily flood forecasts with intelligent data analytic models: multivariate empirical mode decomposition-based modeling methods
Prasad, Ramendra, Charan, Dhrishna, Joseph, Lionel, Nguyen-Huy, Thong, Deo, Ravinesh C. and Singh, Sanjay. 2021. "Daily flood forecasts with intelligent data analytic models: multivariate empirical mode decomposition-based modeling methods." Deo, Ravinesh C., Samui, Pijush, Kisi, Ozgur and Yaseen, Zaher Mundher (ed.) Intelligent data analytics for decision-support systems in hazard mitigation: theory and practice of hazard mitigation. Singapore. Springer. pp. 359-381
A double decomposition-based modelling approach to forecast weekly solar radiation
Prasad, Ramendra, Ali, Mumtaz, Xiang, Yong and Khan, Huma. 2020. "A double decomposition-based modelling approach to forecast weekly solar radiation." Renewable Energy. 152, pp. 9-22. https://doi.org/10.1016/j.renene.2020.01.005
Complete ensemble empirical mode decomposition hybridized with random forest and kernel ridge regression model for monthly rainfall forecasts
Ali, Mumtaz, Prasad, Ramendra, Xiang, Yong and Yaseen, Z.. 2020. "Complete ensemble empirical mode decomposition hybridized with random forest and kernel ridge regression model for monthly rainfall forecasts." Journal of Hydrology. 584, pp. 1-15. https://doi.org/10.1016/j.jhydrol.2020.124647