Prognosis of Diabetes using various Machine Learning Algorithms

Diabetes is considered to be one of the deadliest and chronic diseases that causes a rise in blood sugar. Many difficulties might occur if diabetes stays untreated and unidentified by a specialist. Some of the complications faced due to diabetes are excretory organ injury, typically resulting in chemical analysis, eye damage that may end in visual impairment, or associate degree enhanced risk for cardiopathy or stroke. Usually, the identifying process is tedious which consists of visiting a patient to a diagnostic center and then consulting a doctor. However, the rise in machine learning approaches solves this critical problem. The objective of this paper is to develop a system which can early predict the likelihood of diabetes in patients with maximum accuracy. Hence, machine learning classification algorithms are used in this experiment to detect diabetes at an early stage. Experiments are performed on Pima Diabetes dataset. The performances of various classification algorithms are evaluated on different performance measures such as precision, accuracy, and f-score. Results obtained show Logistic Regression outperforms with the highest accuracy of 82.46% comparatively other algorithms.