LUNG CANCER DETECTION BY HYBRID LEARNING METHOD APPLYING SMOTE TECHNIQUE

Abstract

Lung cancer is a very deadly disease. However, early diagnosis and detection is an essential factor in overcoming this deadly disease. Tumors formed in this disease's initial stage are divided into benign and malignant. These can be visualized using a computed tomography (CT) scan. Thanks to machine learning and deep learning, cancer stages can be detected using these images. In our study, the best and most promising results in the literature were obtained by using a hybrid learning architecture. The data mining techniques we use in obtaining these results also play a significant role. The best accuracy result we obtained belongs to the CNN+GBC hybrid algorithm, which we recommend with 99.71%.