Gated Recurrent Unit and ResNet-50 with Gaussian Naïve Bayes Algorithm Based Infectious Disease Detection

The early detection of infectious disease helps to minimize the spreading of disease with the surrounding environment. Advancement in technology has given automated software for the recognition of infectious disease. The existing methods did not obtain appropriate results for the early detection of infectious disease due to lack of handling the feature dimensionalities. To solve the problem of the existing methods a Gated Recurrent Unit and ResNet-50 with Naive Bayes (GRU and ResNet-50 with NB) is proposed for the infectious disease detection. The chest x-ray dataset is the input data, the normalization and augmentation techniques were used for cleaning the chest x-ray images. The Gated Recurrent Unit (GRU) was used for the taking out of temporal information from the images and further the ResNet-50 extracted the spatial information from chest x-ray image data. The Naive Bayes (NB) is the classifier used for the infectious disease detection. The developed GRU and ResNet-50 with NB has increased the performance of infectious disease detection. The performance of the proposed GRU and ResNet-50 with NB is assessed using the existing metrics such as accuracy 97.89%, recall 97.05%, precision 96.92%, and fl-score 96.98% compared to the existing eXplainable Artificial Intelligence (XAI) method.