CLUSTERING OF DATA POINTS IN MEDICAL DATASETS USING DEEP NEURAL NETWORKS

In healthcare, the classification of medical images plays an important role. The standard approach has nevertheless hit its output ceiling. In addition, time and money must be spent extracting and choosing grouping characteristics by using them. The deep neural network is an innovative form of computer education, which has demonstrated its ability for various classification tasks. In particular, the neural network dominates in varying image recognition tasks with the most successful outcomes. However, it is difficult to gather medical image datasets, so it takes a lot of experience to mark them up. This paper therefore investigates how to identify pneumonia using the convolutional neural network (CNN) on the X-ray chest dataset. To boost output in transfer learning, it is important to retrain special features on a new target dataset. Another important factor is that a good network complexity corresponds to the size of the dataset.