Abstract
This paper presents a new diagnostic model for various diseases. In the proposed diagnostic model, a water wave optimization (WWO) algorithm was implemented for improving the diagnosis accuracy. It was observed that the WWO algorithm suffered from the absence of global best information and premature convergence problems. Therefore in this work, some improvements were proposed to formulate the WWO algorithm as more promising and efficient. The global best information issue was addressed by using an improved solution search equation and the aim of this was to explore the global best optimal solution. Furthermore, a premature convergence problem was rectified by using a decay operator. These improvements were incorporated in the propagation and refraction phases of the WWO algorithm. The proposed algorithm was integrated into a diagnostic model for the analysis of healthcare data. The proposed algorithm aimed to improve the diagnosis accuracy of various diseases. The diverse disease datasets were considered for implementing the performance of the proposed diagnostic model based on accuracy and F-score performance indicators, while the existing techniques were regarded to compare the simulation results. The results confirmed that the WWO-based diagnostic model achieved a higher accuracy rate as compared to existing models/techniques with most disease/healthcare datasets. Therefore, it stated that the proposed diagnostic model is more promising and efficient for the diagnosis of different diseases.
Highlights
In present time, an enormous amount of healthcare data are collected through various sources such as automatic diagnosis system, medical imaging process, and patient information forms like intake, consent, treatment, assessment etc
The performance of the water wave optimization (WWO)-based diagnostic model was assessed over various medical datasets downloaded from the UCI repository
The statistical analysis determines either the simulation results reported the by model/technique are different from other models/techniques or not
Summary
An enormous amount of healthcare data are collected through various sources such as automatic diagnosis system, medical imaging process, and patient information forms like intake, consent, treatment, assessment etc. The abovementioned techniques are widely used for analyzing and developing models for data analysis These algorithms can be either supervised or unsupervised in nature. There is a need of an optimized system for automatic disease diagnosis and patient management that can be built based on computational methods. Numerous clustering algorithms have been developed and categorized based on input data type, similarity measure, type of cluster formed, objective function, and clustering approach (Andreopoulos et al, 2009). This paper presents a WWO-based diagnostic model for diagnosis of different diseases. The proposed model consists of WWO-based clustering technique for determining the labeling of class. The WWO-based clustering algorithm is applied in the diagnosis phase of the proposed model. This algorithm aims to improve the diagnostic accuracy.
Sign-in/Register to view highlights & summary Sign-in/Register to access full text options 
Talk to us
Join us for a 30 min session where you can share your feedback and ask us any queries you have
Schedule a call
