Table 1.

Comparison of Machine learning approaches vs. standard statistical approaches.

	Machine learning	Standard statistics (linear/logistic regressions)
Data preparation?	Doesn't require explicit commands to find patterns in data	Need to know variables and parameters beforehand
Hypothesis?	No hypothesis needed	Need hypothesis to test
Type of data?	Multi-dimensional data that can be non-linear in nature	Linear data
Training?	Needs to be “Trained”	No training
Goal?	Generally better for predictions	Generally better for inferences/hypothesis testing
Scientific question?	What will happen?	How/why it happens?

Table compares the differences between machine learning (ML) and standard statistical approaches (SSA). These comparisons are simplified rules, and therefore may not hold true for every dataset as size and quality of dataset can alter the performances of MLs and SSAs.