Análise de algoritmos de aprendizado de máquina na seleção de atributos para classificação de câncer de mama

Abstract

Cancer is one of the world’s biggest public health challenges and is among the leading causes of mortality. Among its various types, breast cancer stands out for its high incidence, affecting predominantly women, but also a proportion of men. Early diagnosis is essential to increase the chances of a cure. This work aims to carry out a comparative analysis of the performance of different machine learning algorithms applied to breast cancer classification, considering two attribute selection scenarios - a complete set and a reduced set - with the aim of identifying the most effective approaches based on accuracy and precision metrics. The Wisconsin Diagnostic Breast Cancer (WDBC) dataset, with 569 records, was used to train and test different models. The results show that the Random Forest algorithm obtained the best accuracy (96.49%), demonstrating its potential as a medical decision support tool. The proposal aims to help make diagnoses more efficiently and assertively, providing benefits directly to health professionals and patients.