Detection of Cervical Cancer Using CNN and SVM

A remarkable number of cancer-related dead of women in world is causes by cervical cancer. Pap smear examination is most common carry out screening practice for detection of cervical cancer. The study proposes an integrated machine learning approach that integrate convolutional neural network and support vector machine for automating the class of cervical cancer cells. CNN is utilized to extract feature from medical image, their ability to capture spatial hierarchies in the image data and information. The features extracted from the CNN model are then input to the SVM classifier to enhance classification accuracy. The CNN model underwent fine-tuning using a labelled dataset of cervical cell images, while the SVM was employed binary classification distinguishing between normal and abnormal cells. The hybrid CNN-SVM model outperformed both traditional CNN and standalone SVMs, gives higher classification accuracy and robustness.