Pneumonia is a common occurrence in children under the age of five, particularly in those with underlying conditions. It is the leading cause of death in children worldwide. Chest radiography is currently the most effective method for identifying pneumonia. Several research papers have discussed the potential of computer-aided diagnostic tools for pneumonia, but none of them have been designed for pediatric patients. This work addresses the lack of such a tool by proposing a solution based on supervised learning. A comparative study was conducted to ascertain which convolutional neural network performed optimally. The obtained results are comparable with those reported in the literature for adults. The final model is included in a website created for clinical use. The proposed application has the potential to assist clinicians in interpreting data sets and to facilitate accurate diagnoses, particularly in underdeveloped regions where clinicians may lack the necessary expertise.