Paper Title
Earlier Detection and Diagnosis of Lung Cancer Using Convolutional Neural Network

Abstract
Lung cancer is one of the most lethal forms of cancer globally, often diagnosed at an advanced stage due to subtle early symptoms. Early and accurate detection is essential to improve prognosis and reduce mortality. This project introduces a hybrid AI-based framework for lung cancer classification using chest CT scan-derived image data. A Convolutional Neural Network (CNN) built on the pre-trained Xception architecture is employed for robust feature extraction from medical images. The extracted deep features are then classified using multiple machine learning algorithms, including Random Forest, XGBoost, and Support Vector Machine (SVM), enabling comparative performance analysis. The dataset is preprocessed and split into training, validation, and test subsets, ensuring balanced representation across three diagnostic categories: Normal, Benign, and Malignant. Performance is assessed using key metrics such as accuracy, precision, recall, F1-score, and ROC-AUC, with visualization tools like confusion matrices and training-validation accuracy plots aiding interpretability. Fine-tuning of the CNN further enhances prediction capability, and the trained models are exported for scalable deployment. This hybrid approach demonstrates the power of combining deep learning with traditional classifiers in medical image analysis. With the potential for integration into clinical workflows, this system offers a non-invasive, efficient, and interpretable solution for supporting early lung cancer diagnosis. Keywords - Lung Cancer Detection, CNN + XGBoost, Feature Extraction, Xception, Random Forest, Medical Image Analysis, Hybrid Classification