PREDICTING SURVIVAL LIMITATION BY MACHINE LEARNING IN PATIENT WITH CANCER

Authors : Cuma Çakmak, Fadime Çınar, Mehmet Aziz Çakmak

Pages : 842-855

Doi:10.53092/duiibfd.1494646

View : 56 | Download : 124

Publication Date : 2024-11-30

Article Type : Research Paper

Abstract :Cancer is an important public health problem, ranking second in terms of burden of disease in the United States and ranking first in the global burden of disease in the world. Cancer, which causes significant mortality and morbidity, is affected by many factors. Researchers are increasingly interested in this field, both in examining the factors that cause the disease and in managing the disease and are conducting research on this disease with new treatment methods, new techniques and technologies. In this study, its aimed to determine survival rates by analysing open access cancer data representing 8.3% of the US population. With the data obtained, it was tried to classify the survival of cancer patients. Within the scope of the research, various confidence levels were obtained with decision trees, random forrest, SVM algorithms, which are data mining tools. The highest confidence level was obtained with the random forrest algorithm with 75.3%. As a result, it was seen that the model was meaningful and usable, and that survival classification could be made with the data obtained. Survival classification can be an important element for health service providers in resource allocation and effective care.
Keywords : Knime, Veri Madenciliği, Kanser, Ülke Veritabanı, ABD