A Comparative Study of deep Neural Networks for Healthcare Disease Prediction

Abstract- The three deep learning classification models, namely Convolutional Neural Network (CNN), Recurrent Neural
Network (RNN), and Multilayer Perceptron (MLP) are tested to forecast human diseases based on their symptoms in this
study. A publicly available dataset that was gained in Kaggle was involved, and the data were readily used in the classification
models without some further manipulations or feature selections. All the models have been trained on 80:20 training testing
split to provide equal comparison. The standard evaluation measures such as accuracy, precision, recall, and F1-score were
used to evaluate the performance of the classifiers. The experimental findings prove that CNN and RNN models are better
than the MLP model in their ability to classify and generalize. The tested models had the best accuracy of CNN with 85.63 and
then closely the RNN model with 85.16. The results substantiate the usefulness of deep learning methods in prediction of
diseases and emphasize their possible use as a part of smart healthcare.