Cervical cancer risk prediction with robust ensemble and explainable black boxes method

Abstract Clinical decision support systems (CDSS) that make use of algorithms based on intelligent systems, such as machine learning or deep learning, they suffer from the fact often methods used are hard to interpret and difficult understand how some decisions made; opacity methods, sometimes voluntary due problems data privacy techniques protect intellectual property, makes these very complicated. Besides this series problems, results obtained also poor possibility being interpreted; in clinical context therefore it is required accurate possible, transparent explainable results. In work problem development cervical cancer treated, a disease mainly affects female population. order introduce advanced system can be explainable, robust, ensemble method presented, terms error sensitivity linked classification possible aforementioned pathology presented explainability interpretability (Explanaible Machine Learning) applied CDSS Lime Shapley. The obtained, well interesting, understandable implemented treatment type problem.