An analytical study for Pneumonia Detection towards building an intelligent system using Image Data Generator

In India pneumonia is the major reason for mortality in children. To overcome these deaths by pneumonia can be avoided by early detection in time. So that proper diagnosis can be started. In this paper we have described the Convolutional neural network of deep learning classifier as a result we got the result accuracy more than 90%. Different classifier like as support vector machine, k nearest neighbor, random forest and convolutional neural network are helpful for getting the results of the of pneumonia with great efficacy. By usage of the said classifier's pneumonia can be detected at the early stages. CNN with the SoftMax activation function model produces the better precision, specificity but it could not perform well with the radio frequency. At the time of execution CNN produces the best efficacy when we performed the experiment on the different machine learning models. we achieved the accuracy up to 92% with the sequential model. We revised the experimentation to get the maximum accuracy by updating the number of epochs. By this result new hybrid model come up with that accuracy of the more than 90%. Novel classical machine learning approaches produces the better results.