Diabetes is a disease affecting a multitude of people worldwide. Its incidence rates are increasing alarmingly every year. If untreated, diabetes related complications in many vital organs of the body may turn fatal. Early detection of diabetes is very important for timely treatment which can stop the disease from progressing to such complications.
Therefore, the accuracy of diabetes detection and diagnosis has received great attention, and the design and development of a diagnostic system that can successfully identify diabetes at an early stage is a major issue for the scientific community. Existing diagnostic systems have several flaws, including a complex computation and selection of ineffective tools, techniques, or algorithms that affect detected accuracy. So, there is a need for an accurate system for diabetes detection and diagnosis by using the most effective Machin learning techniques to avoid such losses.
In this work, An Automatic System for Diagnosis Diabetes Disease Based on machine learning algorithm was proposed. Furthermore, this work distinguishes itself from the previous by employing Type1 and Type2 Diabetes. dataset with 2000 samples for 8 features. Also, it has the ability to detect and diagnose Diabetes Disease at an early stage. On the other hand, the most effective machine learning techniques for Diabetes disease detection and diagnosis which are KNN, LR, NB, SVM, RF have been implemented.
However, according to the obtained results it is observed that the proposed system achieved an excellent result with accuracy of 99.0% with RF during the comparison with KNN which achieved accuracy of 98.75%, SVM which achieved accuracy of 81.0%, LR which achieved accuracy of 77.50%, and Naive Bayes which achieved accuracy of 77.25%. Also, the performance of the proposed system has been compared with several related works and it is achieved the highest accuracy.