Comparison of PLS-SEM and XAI methods for business-related problems
| Feature | PLS-SEM | XAI |
|---|---|---|
| Developers | Various authors (Ringle et al., 2015; Hair et al., 2019; Henseler, 2020) | Multiple researchers and developers (Adadi and Berrada, 2018) |
| Main focus | Comprehensive PLS-SEM analysis | Interpretability and explainability in AI |
| Interface | User-friendly (e.g. ADANCO, SmartPLS) | Specific to the XAI technique/tool |
| Model evaluation criteria | Incorporates advanced features and algorithms; includes goodness of fit measures | Varies depending on the XAI method |
| Predictive accuracy | Good, but may vary depend on the method and model | Good, but dependent on the particular machine learning model |
| Interpretability and explainability | Fair | Excellent |
| Applicability to complex models | Good | Excellent |
| Prediction, explanation | Balanced emphasis | Explanatory emphasis |
| Feature | PLS-SEM | XAI |
|---|---|---|
| Various authors ( | Multiple researchers and developers ( | |
| Comprehensive PLS-SEM analysis | Interpretability and explainability in AI | |
| User-friendly ( | Specific to the XAI technique/tool | |
| Incorporates advanced features and algorithms; includes goodness of fit measures | Varies depending on the XAI method | |
| Good, but may vary depend on the method and model | Good, but dependent on the particular machine learning model | |
| Fair | Excellent | |
| Good | Excellent | |
| Balanced emphasis | Explanatory emphasis |
Source: This information was compiled by the authors from a comprehensive review of related literature, supplemented with ratings based on the author’s expertise