Table 1 Summary of different artificial intelligence modalities

1. Expert systems [57]

Early AI systems that replicate "expert" decisions. These systems use knowledge bases with organized, factual deductions and heuristics. An expert system learns from these facts and creates "rules" for future decision-making

2. Logistic regression [45]

Predicts binary response variables using a logistic function. It provides simple and reproducible results to compare different complex models

3. Bayesian networks  [38]

These models illustrate variable-outcome connections. They model outcome probability distributions as local, conditional discrete variable probability distributions. It might predict injury risk for an athlete based on current performance measurements and injury history

4. Random Forest algorithms [45]

These algorithms build many "decision trees," flowchart-like structures that emerge from decisions at numerous branching decision points

5. Support vector machines [45]

It creates a multidimensional representation of data as points in space, mapped to distinguish categories as clearly as feasible

6. Artificial and deep neural networks [18]

These models are more independent and require little to no human supervision with less data reformatting