Abstract

Heart disease has become a global health issue and is recorded as one of the primary causes of death in many countries. In this modern era, with rapid technological advancements and shifting lifestyles, numerous factors contribute to the increasing prevalence of heart diseases. These range from dietary habits, lack of physical activity, stress, to genetic factors. Given the complexity of this ailment, information technology plays a crucial role in providing innovative solutions. One of them is predicting the risk of heart disease, enabling more targeted early prevention and treatment interventions.Correct data analysis is pivotal in making predictions. However, a common challenge often encountered is the imbalance in data classes, which can result in a predictive model being biased. This is certainly detrimental, especially in the context of predicting strokes, where prediction accuracy can mean the difference between life and death.In this research, our focus was on developing a Deep Neural Network (DNN) Architecture model. This model aims to offer more accurate predictions by considering data complexities. By optimizing several key parameters, such as the type of optimizer, learning rate, and the number of epochs, we strived to achieve the model's best performance. Specifically, we selected Adagrad as the optimizer, set the learning rate at 0.01, and employed a total of 100 epochs in its training.The results obtained from this research are quite promising. The optimized DNN model displayed an accuracy score of 0.92, precision of 0.92, recall of 0.95, and an f-measure of 0.93. This indicates that with the right approach and meticulous optimization, technology can be a highly valuable tool in combatting heart diseases.

Full Text
Published version (Free)

Talk to us

Join us for a 30 min session where you can share your feedback and ask us any queries you have

Schedule a call