Main Article Content
Stroke is a prominent cause of disability in adults and the elderly, resulting in a slew of social and financial issues. Stroke can be fatal if left untreated. Patients with stroke have been found to have aberrant bio-signals in the majority of cases. Individuals can obtain appropriate therapy more rapidly if they are observed and have their bio-signals detected and precisely assessed in real-time. However, most stroke diagnostic and prediction systems rely on image analysis methods such as CT or MRI, which are costly and difficult to employ for real-time diagnosis. In this paper, we developed a stroke prediction system that detects stroke using real-time bio-signals with machine learning techniques. The purpose of this study was to analyse and diagnose electromyography data using machine learning. The data cleaning was carried out in accordance with the inclusion criteria that were specifically designed. Following that, two data sets were created, each containing 575 facial motor nerve conduction study reports and 233 auditory brainstem response reports. The data sets were then subjected to four machine learning algorithms: Reinforcement Learning, linear regression, support vector machine, and logistic regression. Comparisons of accuracy and recall rate among several algorithms show that the Reinforcement Learning algorithm outperforms the other two in both data sets, Congenital Heart Disease (CHD) and International Stroke Trial (IST). Furthermore, for each algorithm, comparisons were made with and without deviation standardization, and the results show that deviation standardization has an effect on accuracy improvement. The experiment employs three classification algorithms: linear regression, logistic regression, SVM (support vector machine), and Reinforcement Learning. As a result, Reinforcement Learning has been demonstrated to be an optimal algorithm for diagnosis. It is also worth noting that feature ranking in order of importance can facilitate clinical diagnosis and has clinical potential in diagnosis and diagnostic assessment.