| Sumario: | A wide range of Clinical Decision Support Systems (CDSS) have been developed. These CDSS are based on decision models, which normally have a knowledge- or data-driven approach. In this work a structured development of potential hybrid approaches was realized by the assessment of decision models and identification of their advantages and disadvantages. For the assessment of different decision models, eight criteria were identified and three of them were chosen as main criteria for CDSS: transparency, learning aptitude and handling of uncertain and vague knowledge. The comparison of decision models in regard to the developed main criteria resulted in an identification of three groups of models with similar characteristics. Based on these groups hybrid approaches had been developed, so that different decision models could be combined in a beneficial way. Thereby this work provides an instrument for a structured development of hybrid decision models.
|
|---|