Differential incidence of Type 2 Diabetes Mellitus Using Beta-Binomial and Negative Binomial Models. (Research Paper Sample)

International Journal of Scientific Research in ___________________________Research Paper. Mathematical and Statistical Sciences Volume-7, Issue-5, pp.48-52, October (2020)E-ISSN: 2348-4519 Differential Incidence of Type 2 Diabetes Mellitus Using Beta-Binomial and Negative Binomial Models M. Aina1*, T.O. Aro2, D. Tukur3, O.A. Olukiran4, T.A. Oyelakun5 1,3Department of Statistics, Ahmadu Bello University, Zaria, Nigeria 2,4Department of Mathematical and Computing Sciences, KolaDaisi University, Ibadan, Oyo State, Nigeria 5Registry Department, Ladoke Akintola University of Technology, Ogbomoso (LAUTECH), Oyo State, Nigeria *Corresponding Author: taiwo774@gmail.com, Tel.: +2348036176349 Available online at: www.isroset.org Received: 20/Sept/2020, Accepted: 16/Oct/2020, Online: 31/Oct/2020 Abstract—This paper entails, the differential incidence between males and females using Beta-Binomial and Negative Binomial models for type 2 diabetes mellitus. From the previous studies, it was discovered that type 2 diabetes mellitus incidence is at an increased level in the region and localities across the countries. In view of this, gender sensitivity to type 2 diabetes was considered. Secondary data for 10 years were collected from medical records of General Hospital, Ifaki-Ekiti, Ekiti State, Nigeria. The Beta-Binomial and Negative Binomial models were applied to test the incidence of type 2 diabetes in both female and male patients. .The two models were compared with the use of goodness of fit and Akaike information criterion (AIC) for the two fitted models. The experimental results showed that Beta-Binomial outperformed the Negative Binomial model in fitting the incidence of male and female patients in type 2 diabetes. Keywords - Beta-Binomial, Differential Incidence, Negative Binomial, Type 2 Diabetes Mellitus I.INTRODUCTION Diabetes is an ailment triggered by an increase of sugar in the blood as a result of a defect in insulin secretion, insulin action, or the two combined [1]. Patients having diabetes may come up with severe complications such as disease of the heart, failure of kidney, blindness, and early death [2]. The term diabetes mellitus is referred to as a disorder of multiple aetiology categorized by serious hyperglycemia with disturbance of fat, protein, and carbohydrate metabolism as consequent of insulin secretion defection and action [3]. Diabetes mellitus denotes the medical and socio-economic problems in modern society [4] [5]. In developed and developing countries the disease is widespread. It was predictable that the total number of 175 million persons having diabetes in 2004, worldwide and it was further projected that an approximation of 354 million persons will have diabetes by 2030 [6]. The report of the International Diabetes Foundation in 2010, stated that Nigeria is one of the most known countries in Africa that records the huge figure of DM, three million in South Africa, one million and nine hundred thousand in Ethiopia, and one million and four hundred thousand in Kenya 769,000 [7]. Providing a solution for inadequate medical services using the education of human resources requires a long time and high expenses that may result in an increased morbidity rate in patients [8]. This paper conducted a statistical analysis using beta-binomial and negative binomial models to comparatively identify the differential incidence of type 2 diabetes on male and female patients. II. RELATED WORK Many researchers and non-governmental organizations have written a lot about the causes and the avoidance of type 2 diabetes mellitus, in which individuals have come out with their various results and design several models to predict or analyzed various data with different results and recommendations. RoobiniandLakshmi[9]usedmachinelearning techniques to predict diabetes mellitus. Logistic regression, Support Vector Machine and Random Forest were used to perform prediction of diabetes mellitus. A prediction model for the diagnosis of diabetes possibilities early on a patient was developed employing multiple algorithms for improved accuracy in the future. Amira and Nasr [10] proposed a k-means clustering technique for the prediction of diabetes mellitus. Data were pre-processed to remove data inconsistency or uncertainty and noise from the dataset through the implementation of a clustering model. The study applied the k-means approach to community health diabetes-related indicators dataset to cluster diabetic patients from healthy ones with high accuracy and reliability results. Kuyoro, Osisanwo and Franlyln [11] developed an online diagnostic System for managing early diabetes neuropathy. Four techniques of classification; KNN, C4.5, Decision ...