A Comparative Study of Machine Learning Algorithms using Quick-Witted Diabetic Prevention

Abstract

Wireless Sensor Networks (WSN) are the latest advances in major information technology. Cutting-edge multi-Gbps data accelerates networks and large-scale medical data analytics have enabled the development and implementation of new diabetes control systems applications. Advanced multi-Gbps data speeds allow a new type of network designed to connect virtually including machines, objects, and devices. It is important to develop effective methods for the diagnosis and treatment of diabetes because of its long-term and systemic effects on diabetic patients. The current diabetes screening system faces the following problems: the system is inconvenient and inconsistent, difficult to collect real-time data and predicting the disease severity based on the food habits of the patients which changes day today. The diabetes screening model lacks data-sharing mechanisms, patient behavior, and personal testing. As a result, there are no continuing proposals for the prevention and treatment of diabetes. To resolve these problems, proposed a next-generation diabetes solution called the Smart Diabetes Diagnosis System. This technology used to build this system includes machine learning, massive medical data, and a large cloud of health intelligence. An intelligent diabetes management system can provide a clearer and more focused personal risk assessment and treatment schedule, and provide patients with detailed daily guidance on how to improve their self-medication. A test model has been developed to test the feasibility of multiGbps peak data speeds.  Machine Learning algorithms to assure the performance of our Smart Diabetes Test – Decision Tree, Support Vector Machine (SVM), Artificial Neural Networks (ANN), Naive Bayes approved. This paper has attempted to demonstrate that our system can provide patients with individual diagnoses and treatment recommendations. Smart Diabetes System utilizes advanced data classification and tracing techniques to visualize patient monitoring graphics to improve recovery rate periodically.