LCDNET-Lung Cancer Detection Using Deep Neural Network

In many countries lung cancer is acknowledged mostly as a deadly illness; it is recommended that the illness should be diagnosed early. Different medical Population Based information sources are employed in the analysis to make predictions of the related risk of lung cancer using different learning algorithms. DNN approach out performs CNN and many other approach because of the depth in the architecture of the deep neural network. All the features are preprocessed where the categorical variable is converted into numerical and then the data is prepared where necessary to enhance the model effectiveness. The dropout and L2 regularization incorporated into the DNN models reduce the overfitting and increase the model's generalization capacity. The metrics, accuracy, precision recalls, and F1-score which are utilized in the proposed model by to better other models' results with 98% accuracy. 54%. This study also reveals the feature of the deep learning models to fast diagnosis of the lung cancer meaning that the cancer patients would be given the right treatments. In this regard, multiple considerations on the basis of Advanced Deep Learning methods may reduce the global effects of lung cancer-related diseases. The future serious works will progress to the model by adding more clinical features and to examine the model with larger and more miscellaneous samples.