Pneumonia identification can be facilitated through the application of Machine Learning models. We worked with a Kagle dataset of 5,856 X-ray images comprising both normal and pneumonia cases. To ensure balanced representation, the dataset was rebalanced with an 80-10-10 split for training, testing, and validation respectively. The evaluation of the models' performance relies on two commonly used performance metrics: recall and accuracy. Motivated by the need for more reliable diagnostic criteria, this paper reviews various computer vision methods and identifies the Convolutional Neural Network (CNN) and Logistic Regression approaches as the most promising techniques for this task. After implementing these two models (CNN and Logistic Regression), conducting hyperparameter tuning, and testing various data transformations, the results of eight models show that logistic regression is most effective at predicting pneumonia with a recall of 99.4%. Additional research can explore if this may be due to sample size, number of epochs, or the success of certain image augmentations.