Cervical cancer is a leading cause of cancer-related deaths among women worldwide. Early detection and accurate classification of cervical cancer play a crucial role in improving patient outcomes. In this work, we propose a novel approach for cervical cancer classification that leverages multi-scale feature extraction using Convolutional Neural Networks (CNN) and classification via extreme Gradient Boosting (XG-Boost). The approach first applies a multi-scale CNN architecture to extract hierarchical and spatial features from cervical images, capturing both fine-grained and large-scale structural information. These extracted features are then fed into an XG-Boost classifier, which is optimized to make accurate predictions based on the extracted multi-scale representations. The proposed method is evaluated on a publicly available dataset, and its performance is compared with other state-of-the-art classification techniques. Our results show that the combination of multi-scale feature extraction with CNNs and the powerful boosting capability of XG-Boost significantly improves the accuracy and robustness of cervical cancer classification. This approach demonstrates potential for real-time applications in clinical decision-making systems, enhancing the early detection of cervical cancer and ultimately improving patient care.