A Revisited Convex Hull-Based Fuzzy Linear Regression Model for Dynamic Fuzzy Healthcare Data

Abstract

Healthcare data analysis is widely used in cancer classification and disease prediction. Hence, fuzzy linear regression integration into dynamic analysis can provide better decision making process in healthcare industry especially dealing with dynamic data analysis. Healthcare officers and related researchers require efficient regression tools to produce precise inference as an aid to save human life. However, the key problem in this circumstance is related to computational complexity and processing time. Both of these parameters drastically increase beside an increment of data size in dynamic databases. With regard to the aforementioned problem, the main objective is to improve the implementation of fuzzy linear regression for fuzzy data by addressing and mitigating some limitations of the existing methods through a convex hull approach. More specifically, we look at the realization of an incremental algorithm called Beneath-Beyond algorithm. This algorithm provides a useful vehicle to reduce the computing time and computational complexity as well. Furthermore, a real fuzzy healthcare data set which derive from healthcare industry will be selected as the main source of data sets. Additionally, there are two major procedures or components namely a formulation of a product of fuzzy number optimization and the use of the convex hull technique to the obtained locus points in hyper-rectangles polygon and each of them have their own distinctive activities. As a research output, the combination of this mathematical geometry algorithm and fuzzy linear regression analysis will produce an optimized algorithm called convex hull-based fuzzy linear regression model deliberately for dynamic fuzzy healthcare data. The proposed algorithm may help to produce a rapid decision making especially for critical area such as healthcare industry.