Prediction of Onset of Diabetes using Adaptive Boosting
Pushpa S. K1, Manjunath T N2, Bhavya G3, Vinutha K4, Anupchandra Rao M C5

1Pushpa S. K, ISE department, BMSIT&M, Bengaluru, India.
2Manjunath T N, ISE department, BMSIT&M, Bengaluru, India.
3Bhavya G, ISE department, BMSIT&M, Bengaluru, India.
4Vinutha K, ISE department, BMSIT&M, Bengaluru, India.
5Anupchandra Rao M C, ISE department, BMSIT&M, Bengaluru, India.
Manuscript received on January 02, 2020. | Revised Manuscript received on January 15, 2020. | Manuscript published on January 30, 2020. | PP: 1371-1376 | Volume-8 Issue-5, January 2020. | Retrieval Number: E6303018520/2020©BEIESP | DOI: 10.35940/ijrte.E6303.018520

Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (

Abstract: Diabetes is one of the most common diseases, as per the survey in 2015, 30 million people in US are suffering from this disease, i.e about 90-95 percent of the population. If diabetes is untreated at the early stages, high blood glucose in the body leads to various other health problems like: eye problems, stroke, nerve damage, heart disease, stroke etc. Technology has seen an explosive growth in the development and use of Artificial Intelligence in various domains. The increased sophistication and capabilities of these tools are unlocking new possibilities in fields of Medicine, Agriculture, Manufacturing and Automobiles. The goal of this work is to predict the onset of diabetes using Machine Learning namely Adaptive Boosting. Boosting is a technique wherein a series of low accuracy classifiers are combined to create a high accuracy classifier. In many areas the problems are so complicated that simple algorithms such as KNN, Decision Tress are incapable of making predictions. Hybrid algorithms such as Random Forests and Gradient Boosting are gaining popularity due to these reasons are used by multinational companies one example being Netflix. In this work Decision Tree and Support Vector Machine methods has been considered with eight important attributes namely, Pregnancies, Glucose, Blood Pressure, Skin Thickness, Insulin, BMI, Diabetes Pedigree Function, Age and predicts if a person has diabetes. Multiple models are built using decision tree and support vector machine without Adaptive Boosting and with Boosting technique and the results are compared and evaluated. Result shows that support vector machine gives an improved overall accuracy of 80%.
Keywords: Machine Learning, Ensemble Learning, Ada boaster.
Scope of the Article: Machine Learning.