Entropy is a very wide-meaning scientific term. This article specifically associates “entropy” with the state of disorder, randomness, chaos, and uncertainty in data science analysis perspectives. More specifically, machine learning (ML) and data mining (DM) mathematical models are the environments in which entropy would be related and referenced in its contexts. What are the key roles of entropy in ML and DM? How does it affect results and interpretation procedures? Are the currently used models considering the entropy and aware of major mathematical faults, in incorrect or improper entropy prediction? Lastly, what are the implications in such a case? The article is based on a few years of research, for a Ph.D. thesis, theoretical and practical research field. The article suggests practical implemental software-based apparatus to increase precision, accuracy, and possibly better math-models results after performing some math-models analysis. It offers a simple pre-process stage which decreases the possible faults of wrong entropy pre-process use.