LUNG CANCER DISEASE PREDICTION USING MACHINE LEARNING

Abstract

 Lung cancer is still a global health problem in recent years because it has taken lives of millions of people. Diagnosis for lung cancer relies on manual interpretation of CT scans which causes a great deal of variability in diagnosis of lung cancer. In this paper, we proposed advanced hybrid model that perform contrast enhancement of CT scan images using histogram equalization, extract the features using Densenet121 Deep learning (DL) methods for automated lung cancer analysis based on CT scan images. These features fed into the stacking model for the classification. This stacking model uses Logistic Regression (LR), Support vector machine (SVM), Random Forest (RF), XGBoost and K-Nearest Neighbour (KNN) and base model and implemented Logistic regression as base learner.  The result of the proposed methodology demonstrates accuracy of 98.5% with a precision of 98.2%, recall of 97.8%, and F1-score of 98.4%.  The existing systems uses standalone Convolutional neural network (CNN), which and other transfer learning Deep learning model for classification of CT scan images. The proposed model outperforms than the existing model and shows robustness.