High Performance of Predict Epileptic Seizures System Based on Machine Learning with IoT.

This paper proposes a simulation model that enhances monitoring and warning systems, providing significant assistance to persons with epilepsy and their careers. Smartphones may be categorized as Internet of Things (IoT) devices that fulfill data processing and communication functionalities across both private and public domains. The next iteration of smartphones incorporates sensor systems. The analysis of motion shown by mobile objects may be ascertained via the use of accelerometers and other integrated sensors inside these components. Accelerometers has the capability to identify many physiological occurrences, such as epileptic convulsions and unexpected assaults. A significant proportion of patients diagnosed with epilepsy have epileptic episodes, resulting in impaired equilibrium and potential harm or physical impairment. The data obtained from the acceleration sensor included in the smartphone was used in both the construction and subsequent assessment of the Epilepsy alert system. The data was obtained by collecting measurements from three acceleration sensors that were implanted into mobile phones, using the MATLAB program. A sample size of 20 participants was chosen to serve as representatives of persons diagnosed with epilepsy across 13 unique settings. Furthermore, a total of 30 people were chosen as datasets, whilst 6 individuals were allocated for the purpose of testing the system. The MATLAB programming language was used to compute speeds and distances as a means of evaluating and presenting the outcomes of data processing for smartphones. The use of automated correlation analysis between the movement of the patient and the reference data was employed as a means to determine if the observed movement exhibits the qualities of normalcy or abnormality. The described technique for detecting epilepsy had a matching rate of up to 85 percent.