Effective Diagnosis of Diabetes with a Decision Tree-Initialised Neuro-fuzzy Approach
Diabetes mellitus is a serious hazard to human health that can result in a number of severe complications. Early diagnosis and treatment is of significant importance to patients for the acquisition of a better quality life and precaution against subsequent complications. This paper proposes an approach by learning a fuzzy rule base for the effective diagnosis of diabetes mellitus. In particular, the proposed approach starts with the generation of a crisp rule base through a decision tree learning mechanism, which is data-driven and able to learn simple rule structures. The crisp rule base is then transformed into a fuzzy rule base, which forms the input to the powerful neuro-fuzzy framework of ANFIS, further optimising the parameters of both rule antecedents and consequents. Experimental study on the well-known Pima Indian diabetes data set is provided to demonstrate the promising potential of the proposed approach.
- 1.Holt, R.I., Hanley, N.A.: Essential Endocrinology and Diabetes, vol. 41. Wiley, Chichester (2012)Google Scholar
- 15.Breiman, L.: Classification and Regression Trees. Routledge, New York (2017)Google Scholar
- 17.Bache, K., Lichman, M.: UCI Machine Learning Repository (2013). http://archive.ics.uci.edu/ml