[This article belongs to Volume - 56, Issue - 01, 2024]
Gongcheng Kexue Yu Jishu/Advanced Engineering Science
Journal ID : AES-27-05-2024-51

Subhash Chandra Jat, Mona Nagar

Abstract :

Cancer is the most serious illness to human beings. Previous attempts to evaluate cancer illnesses assisted in establishing which progressions should be conducted on high-risk patients, lowering their risks. These cells may eventually cause cervical cancer. Cervical cancer is the top cause of death among women, and early identification is essential for successful treatment. Recent research has looked into the use of ML for early diagnosis of cervical cancer, but problems remain. This study compares the performance of various MLtechniques, like XGBoost, DT, AdaBoost, CatBoost, and a Stacking model, in forecasting cervical cancer. A dataset for cervical cancer prediction was used from the UCI ML Repository in the proposed project. Data preprocessing, which involves fixing data imbalances and extracting features, begins with splitting the dataset into a training set and a testing set. The predictive system's performance is assessed using classification accuracy, F1-measure, recall, precision, confusion matrix, and classification report. The suggested model detects cancer among patients using multiple ML classifiers, with a Stacking model having a greatest accuracy of 0.9687%. A results demonstrate that asuggested method achieves better results than standalone ML algorithms on many measures, including projected accuracy. The report emphasises the need of further research in this field, as well as the potential of ML to improve cervical cancer diagnosis. The research intends to contribute to efforts to enhance cervical cancer diagnosis and treatment by offering a novel strategy that solves the problems faced by existing approaches.