Cardiovascular disease has been the most common cause of death for the last few decades in India and worldwide. Optimal prediction of cardiovascular disease is necessary for efficiently treating patients before the occurrence of the disease. Several intelligent healthcare frameworks have been proposed for the prediction of cardiovascular disease, employing machine learning and deep learning approaches. However, there is still a need for an enhanced and optimal predictive system. This paper introduces a deep feed-forward neural network (DFFN) approach and regularization technique that is early stopping for classification and overcomes the overfitting problem to enhance accuracy in a shorter amount of training time. Utilizing a benchmarked dataset, this optimal model achieved 98.58% accuracy in predicting cardiovascular disease, and this is contrasted with various machine learning algorithms, including logistic regression, support vector machines, Naive Bayes, random forests, and artificial neural networks.