Diagnosing leukemia is one of the most difficult issues facing hospitals and medical teams due to the need to make quick medical decisions in a short time before treatment of this disease becomes impossible especially for children because they are considered the ones most affected by leukemia. Fortunately, there are many scientistic sides that made a lot of studies and theories for serving the medical domain, especially artificial intelligence. This research introduces a technique to classify microarray data analysis with machine learning to detect leukemia using imaging processing. The data has been taken from both infected and non-infected people in laboratory image format. The classification algorithms which are used in this research SVM, KNN, and Naive Bayes. The main goal for this research is to provide the most accurate percentage among these algorithms and therefore that leads to providing a high prediction of leukemia.