A Study of the Accuracy of the Model Sequential in Deep Learning: Focusing on a Survey

Abstract

The ultimate goal of this paper is to provide an in-depth analysis of the accuracy of the model sequential in deep learning. A point to note is that the sequential model worked properly for train data and validation data and that this took place after epoch 4. When learning took place once, the accuracy rate of the model sequential was considerably low, whereas learning happened twice, that of the model sequential improved dramaticaly. A further point to note is that after epoch 5, train data and validation data were perfectly predicted by the sequential model. A major point of this paper is that the val_accuracy of the sequential model reached a peak from the beggining to the end, whereas the accuracy of the sequential model increased dramatically to 100% after epoch 3. It is worthwhile pointing out that the sequential model judged cold patients as cold patients, whereas it judged no cold patients as no cold patients. This in turn indicates that test data were perfectly predicted by the sequential model. It must be stressed, on the other hand, that the proportion of the so-called recall is 100%. Put differently, our model judged true as true. It therefore seems reasonable to conclude that test data were also perfectly predicted by the sequential model, hence indicating that the sequential model counts as a good model.