Cancer Prediction Using Random Forest and Deep Learning Techniques

Cancer is a medical condition in which cells divide abnormally in the human body and spread to other body parts. If the prognosis of cancer is done well in advance, it is curable. It is estimated that 22.2 million individuals would be impacted by cancer between now and 2030. There are about 100 different forms of cancer, which are categorized based on criteria including age, gender, and race/ethnicity. Cancer in organs such as the lung, liver, kidney, breast, and brain was investigated in the present study. The data for the Lung Cancer study came from the Kaggle online system, the TCIA informed on kidney cancer (The Cancer Imaging Archive), the HCC Liver Cancer dataset, the Brain Cancer dataset from data. World, and the Wisconsin dataset for breast cancer. The proposed approaches to classify cancer's malignancy were Random Forest Classifier, Convolutional Neural Network, and ResNet50. Lung, Liver, Kidney, Breast, and Brain cancers had accuracy rates of 99 percent, 75.75 percent, 88.09 percent, 96 percent, and 81 percent, respectively.