Fuzzy-ID3 Algorithm for Breast Cancer Disease Classification

In terms of death rates, breast cancer comes in second, among women with cancer. Despite the fact that cancer cells grow in a multistep process involving a number of different types of cells, prevention of breast cancer stays a challenge inside the modern world. As a method of breast cancer detection, this paper proposes ENTROPYMOC strategy, a fuzzy decision tree with a new formula of Entropy. It aims to improve the classification accuracy, precision, recall and F1-Measure of the decision tree by overcoming the limitations of the ID3 algorithm, which is not able to classify continuous-valued data. In the field of machine learning, fuzzy decision trees are becoming increasingly popular. This algorithm reduces the complexity of the logarithmic entropy formula by simplifying the Shannon entropy principle. WBCD (Original), WDBC (Diagnostic) and Coimbra datasets are used to test the improved algorithm. Based on the experimental results, the improved fuzzy-ID3 algorithm outperforms the other four classification algorithms (SVM, Naïve Bayes, Random forest and FId3) in terms of accuracy. In Coimbra dataset, accuracy increased by 3.448%.