CBR Outcome Evaluation for High Similar Cases: A Preliminary Approach

* Final gross prices may vary according to local VAT.

Get Access

Abstract

Case-based reasoning has demonstrated to be a suitable similarity-based approach to develop decision-support system in different domains. However, in certain scenarios CBR finds difficulties to obtain a reliable solution when retrieved cases are highly similar. For example, patients from an Intensive Care Unit are critical patients in which slight variations of monitored parameters have a deep impact on the patient severity evaluation. In this scenario, it seems necessary to extend the system outcome in order to indicate the reliance of the solution obtained. Main efforts in the literature for CBR evaluation focus on case retrieval (i.e. similarity) or a retrospective analysis. However, these approaches do not seem to suffice when cases are very close. To this end, we propose three techniques to obtain a reliance solution degree, one based on case retrieval and two based on case adaptation. We also show the capacities of this proposal in a medical problem.

This study was partially financed by the Spanish MEC through projects TIN2006-15460-C04-01, PET2007_0033, the SENECA 08853/PI/08, and the Excellence Project P07-SEJ-03214.