Văn bản này: How fair can we go in machine learning? Assessing the boundaries of accuracy and fairness