Early detection is crucial in improving lung cancer outcomes. However, lung cancer often remains asymptomatic in its early stages, leading to late-stage diagnoses in many cases. Computed Tomography (CT) scanning has emerged as a powerful tool for lung cancer screening and diagnosis, offering detailed, three-dimensional images of the lungs that can reveal small nodules or tumors before they become symptomatic. The interpretation of CT images, however, is a complex and time-consuming task that requires significant expertise. Radiologists must carefully analyze numerous images to identify potential malignancies, a process that is susceptible to human error and fatigue. This challenge has spurred the development of computer-aided detection (CAD) systems, which aim to assist radiologists by automatically identifying and classifying suspicious areas in CT scans. Recent advancements in artificial intelligence and machine learning have opened new avenues for improving the accuracy and efficiency of lung cancer detection and classification. These technologies offer the potential to analyze vast amounts of imaging data, recognize subtle patterns, and provide rapid, consistent results. By augmenting human expertise with machine learning algorithms, we can potentially enhance early detection rates, reduce false positives and negatives, and ultimately improve patient outcomes. This study investigates the effectiveness of various machine learning models for lung cancer detection and classification using CT images. The research employs Auto Color Correlogram (ACC) features and compares the performance of six classifiers: Additive Regression (AR), Naive Bayes (NB), Linear Regression (LR), Attribute Selected Classifier (ASC), Naive Bayes Multinomial (NBM), and Logistic Regression. The results demonstrate that the AR model outperforms other classifiers across multiple evaluation metrics. AR achieves the highest accuracy at 90.10%, precision of 0.90, recall of 0.89, ROC of 0.97, and PRC of 0.97. It also exhibits superior performance in terms of kappa statistic (0.65), F-Measure (0.89), and Matthews Correlation Coefficient (0.65). In contrast, the ASC model generally shows the lowest performance, with an accuracy of 84.10%, precision of 0.81, recall of 0.83, ROC of 0.84, and PRC of 0.78. The ASC model also has the lowest kappa value (0.53) and Matthews Correlation Coefficient (0.53) among the compared models. Notably, the Logistic Regression model matches the AR model in precision (0.90), while the NBM model shows the lowest F-Measure (0.34) among all classifiers. These findings suggest that the Additive Regression model, when combined with ACC features, offers a promising approach for automated lung cancer detection and classification in CT images. This research contributes to the ongoing efforts to enhance computer-aided diagnosis systems in oncology, potentially improving early detection and classification of lung cancer.