Python implementation to the paper "Evaluation of computationally intelligent techniques for breast cancer diagnosis" published on Neural Computing and Aplications journal (2021).
Paper seems to be using ID attribute in the dataset to train AI agents, but it does not make sense since ID is only a sequence. So we need to check if they really use it or just forgot to inform in their paper, based on our results.
Paper calculates the confusion matrix over the entire dataset (see section 6.1), I don't think it's correct, but let's implement it as they did since our goal is to reimplement the code. It is important to define the variables TP, FP, TN and FN (vide section 4) since they are used as basis to calculate the metrics defined in 4.1-4.9. Besides the definition of variables, it should show a table with the confusion matrix values.
Use the metric values obtained for each classifier type and generate plots similar to figures 16 and 17. Remember that you should generate an adaptative code that works with many classifiers without require code modifications.