Hierarchy Representation of Data in Machine Learnings

Machine learning is currently the subject of extensive research efforts aimed at enhancing its performance. Research that exclusively emphasizes an empiricist perspective, viewing all knowledge as derived from empirical experiences, has enjoyed a great success. However, for effective model improvement, it is essential not only to focus on the model but also to investigate the data. Balancing the rational perspective, which estimates models based on datasets, is essential for improving the model's learning process([8]). Additionally, it is argued that understanding the cause-and-effect perspective, independent of how data fits, is vital, and acquiring such a perspective is crucial([9]).