As the Internet of Things becomes more prevalent in our lives, we are confronted with more security concerns. Network attacks have become more common in the cyber world these days. Denial of service, Prove, Remote to Local attacks, and other types of attacks are on the rise. In our research, we used five machine learning classifiers and conducted a comparison analysis to see which one performed better in predicting network anomalies. Since the dataset we used is unbalanced, we experimented with oversampling and under sampling techniques for the minority and majority groups to improve the model's prediction. Then, in order to test and compare our models, we measured accuracy, F1 ranking, and confusion matrix.