Identification of childhood leukemia using deep learning

This thesis report is submitted in partial fulfillment of the requirements for the degree of Bachelor of Science in Computer Science and Engineering, 2017.Cataloged from PDF version of thesis report.Includes bibliographical references (page 28).Although cancer in children is rare, it is the leading cause of death past infancy amongst children. According to Afshar, Abdolrahmani, Tanha, Seif, Taheri(2010), Leukemia or blood cancer is one of the most common cancers in children, comprising of more than a third of all childhood cancers. Despite the advances of technology and research and overall decrease in mortality, nearly 2000 children die of cancer each year in the United States according to www.cancer.gov(2017). The website also tells us that if Leukemia cases are identified late or proper treatment isn’t applied, then it can be mortal. For this reason, we have decided to use deep learning for the rapid identification of leukemia in the absence of doctors, which can be done in clinics by present nurses and lab workers. We are going to use ID3 and C4.5 (extension of ID3) classifiers, Naïve Bayes and Multi-layer Perceptron (MLP) Neural network on the data I have gathered of the 78 cases and check which one gives the most accurate result.Farana Naz TultulB. Computer Science and Engineerin