Literature review

Literature review

Different articles have surveyed the literature that has been published on varying prediction models through the use of data mining to diagnose and predict diabetes. Diabetes has been considered as one of the severe disease amongst the well acknowledged non-communicable diseases in the universe. The objective of the following related literatures is to cultivate a extrapolative healthcare result support structure through the use of classifications and data mining techniques.

According to this Wu et al., (2018) there has been increased occurrence of families being affected by Diabetes mellitus. The study projected a novel reliant on data mining methods to enable predict the type 2 diabetes mellitus. The chief problem that the article is attempting to resolve is the improvement of accuracy of the forecast model and to ensure that the technique is more effective than one data set. The model is useful in the management of the disease with the model being able to attain 3.04 percent higher accuracy as compared to other researchers.

With regards to the need of predictive techniques for diabetes Mirza and Mittal and Zaman (2018) came up with a decision supportive model for diabetes prognosis using SMOTE and Decision tree through the use of decision tree classifier and SMOTE. In this case the classification of imbalance info especially in medical informatics it becomes a bit challenging. This was then the main motivator for the development of SMOTE. In the second stage diabetes was diagnosed through the use of decision tree classifier. The acquired classification accuracy is 94.7013 which was better compared to other methods relied.

Relatively Hashi, Zaman and Hasan (2017) described the clinical support decision system which relies on the perception and experience of the doctor rather than relying on complex statistical data. The conventional disease diagnosis system proposed aims at assisting doctors in predicting the disease correctly which benefits the patient and the medical insurance. The system in the case of diabetes has relied on the k-Nearest Neighbor Algorithms and the Decision Tree which as classification techniques. The article reveals that the C4.5 offers better accuracy compared to the KNN for diabetes diagnosis. The aim of Maniruzzaman et al., 2018 is the development of a robust and optimized machine learning system under assumptions of the outliers or the missing values if substituted by the median configuration which results greater risk of stratification accurateness. The R-F based model showed better performance when the omitted values and outliers were substituted by the median values.

Rubaiat, Rahman and Hasan (2018) introduced an approach through which diabetes diagnosis can be approached through doctors automatically predicting the type 2 diabetes mellitus through relying on the neural network. Analysis was done through input the features on the MLP neural network classifier which acquired a precision of 85.15 percent. The conclusion was that prediction of diabetes through MLP neural network were far better as compared to the K-means.

On the other hand, Zhu, Idemudia and Feng (2019) aimed at determine ways through which K-means can be improved to offer better results. The model comprised of principal component analysis, logistics regression and K-means clustering. The results revealed that enhanced K-means and logistic regression through the PCA offer a higher accuracy of 1.98%. This model shows effective prediction of diabetes through reliance on electronic health data records.

References

Hashi, E. K., Zaman, M. S. U., & Hasan, M. R. (2017, February). An expert clinical decision support system to predict disease using classification techniques. In 2017 International Conference on Electrical, Computer and Communication Engineering (ECCE) (pp. 396-400). IEEE.

Maniruzzaman, M., Rahman, M. J., Al-MehediHasan, M., Suri, H. S., Abedin, M. M., El-Baz, A., & Suri, J. S. (2018). Accurate diabetes risk stratification using machine learning: role of missing value and outliers. Journal of medical systems, 42(5), 92.

Mirza, S., Mittal, S., & Zaman, M. (2018). Decision Support Predictive model for prognosis of diabetes using SMOTE and Decision tree. International Journal of Applied Engineering Research, 13(11), 9277-9282.

Rubaiat, S. Y., Rahman, M. M., & Hasan, M. K. (2018, December). Important Feature Selection & Accuracy Comparisons of Different Machine Learning Models for Early Diabetes Detection. In 2018 International Conference on Innovation in Engineering and Technology (ICIET) (pp. 1-6). IEEE.

Wu, H., Yang, S., Huang, Z., He, J., & Wang, X. (2018). Type 2 diabetes mellitus prediction model based on data mining. Informatics in Medicine Unlocked, 10, 100-107.

Zhu, C., Idemudia, C. U., & Feng, W. (2019). Improved logistic regression model for diabetes prediction by integrating PCA and K-means techniques. Informatics in Medicine Unlocked, 100179.