Human integration in AI : calibrating trust and improving performance in decision support systems / Jaroslaw Kornowicz, M.Sc. Paderborn, 2025
Inhalt
- Contents
- List of Figures
- List of Tables
- Synopsis
- The Role of Response Time for Algorithm Aversion
- Algorithm, Expert, or Both? Evaluating the Role of Feature Selection Methods on User Preferences and Reliance
- Introduction
- Related Work
- Methods
- Participants and Treatments
- Experimental Procedure
- Classification Domains and Machine Learning Models
- Evaluation Measures
- Results
- Discussion
- Conclusion
- Appendix
- An Empirical Examination of the Evaluative AI Framework
- Introduction
- Background and Related Work
- Method
- Overview
- Experimental Conditions
- Hypotheses and Dependent Variables
- Decision-Making Process
- Procedure
- Dataset
- Regression Model for Income Assessment
- Participants
- Results
- Discussion
- Limitations and Future Work
- Conclusion
- Appendix
- Towards a Computational Architecture for Co-Constructive Explainable Systems
- Aggregating Human Domain Knowledge for Feature Ranking
- Introduction
- Related Work
- Feature Ranking and Selection
- Expert Judgement Elicitation
- Combining Feature Ranking with Expert Judgement Elicitation
- Materials and Methods
- Results
- Discussion
- Conclusion
- Appendix
- Comparing Humans and Algorithms in Feature Ranking: A Case-Study in the Medical Domain
- Introduction
- Data- and Knowledge-Driven Feature Selection
- Probabilistic Scoring Lists
- Evaluation
- Conclusion
- Human-AI Co-Construction of Interpretable Predictive Models: The Case of Scoring Systems
- Introduction
- Scoring Systems and Extensions
- Co-constructive Framework: PSLvis
- Method
- Results
- Participants' Assessments
- Co-construction as Navigation in the Model Space
- PSLvis User Actions
- Think-Aloud Results
- Co-Construction with PLSvis
- Decision-Making with PSL
- Discussion
- Limitations and Future Research
- Appendix
- Bibliography