Machine Learning based Algorithmic approach for Detection and Classification of Leukemia

Abstract

In general, leukemia is diagnosed by taking repeated complete blood counts, since this will enormously increase the blood cell count of the patient compared to normal people. The malignant cells resemble the normal blood cell which complicates the prediction process. So, this aliment must be detected and treated in early stages to avoid any complications. The methods already existing in laboratories are time consuming. This study presents a Machine Learning approach for detecting leukemia in patients. A dataset consisting of blood smear images was collected and preprocessed to extract relevant features. The features were then used to train and test various machine learning classification algorithms such as Logistic Regression, Decision Tree, Random Forest, Support Vector Machine and KNN. Then by evaluating the performance of the above-mentioned classifiers using different performance metrics like accuracy, precision, etc., the efficient one can be identified.