Analysis of neuronal signals may reveal the presence of epilepsy, a severe persistent neurological condition. In order to transmit impulses and interact with bodily functions, neurons are intricately linked to one another. Electroencephalography (EEG) and electrocorticography (ECoG) are popular tools for monitoring these brain impulses. These signals generate a great deal of data, are complicated, noisy, non-linear, and non- stationary. On a global scale, some 50 million individuals have epilepsy, with 100 million being impacted at some point throughout their lives. The stated prevalence rate is between half a percent and one percent, and it is responsible for one percent of the global illness burden. Multiple seizures occurring in a given patient is the hallmark of epilepsy. It triggers an abnormal electrical storm in the brain, which in turn causes the patient to experience a transient loss of consciousness, altered behavior, and altered sensations. As a result, finding new information about the brain and detecting seizures are both difficult tasks. Hence, the suggested approach offers a practical means of predicting the identification of epileptic seizures by integrating brain and heart signals more effectively. In this study, we will use ensemble models of machine learning algorithms, including logistic regression, random forest, and support vector machines (SVMs), to detect seizures early on. After gathering datasets from the physio net dataset, we will train them using a machine learning technique and then build the model file. It is possible to accurately detect the existence of epileptic seizures and encourage successful treatment when input data is used for seizure prediction. With a combined brain–heart signal that is more accurate than current models, this study aids in the successful diagnosis of seizure determination.