Related Experiment Videos
Representation of preferences in decision-support systems
1Section on Medical Informatics, Stanford University.
Computers and Biomedical Research, an International Journal
|August 1, 1992
Summary
Decision-theoretic preference models enhance computer-based decision-support systems by explicitly capturing expert preferences. This study presents an efficient method to automatically refine these models using simulated expert decisions.
Area of Science:
- Computer Science
- Decision Analysis
- Artificial Intelligence
Background:
- Computer-based decision-support systems (CDSS) often rely on implicitly represented expert preferences.
- Explicitly modeling decision-maker preferences offers significant advantages for CDSS functionality and reliability.
Purpose of the Study:
- To describe the advantages of explicit decision-theoretic preference models in CDSS.
- To present an accurate and efficient method for determining and refining these preference models.
Main Methods:
- Simulating familiar decisions within an expert's domain of expertise.
- Inferring expert preferences based on choices made in simulated decision scenarios.
- Automatically refining the decision-theoretic preference model using inferred preferences.
Main Results:
- The proposed method accurately and efficiently determines expert preferences.
- The automated refinement process improves the fidelity of the preference model.
- Explicit preference models enhance the performance and applicability of CDSS.
Conclusions:
- Explicit decision-theoretic preference models are crucial for effective CDSS.
- The developed preference assessment method offers a practical solution for model creation and refinement.
- This approach facilitates the development of more personalized and reliable decision-support tools.