Intelligent Signal Analysis Using Case-Based Reasoning for Decision Support in Stress Management - Computational Intelligence in Healthcare: Advanced Methodologies - page 157

New Case

Problem

Learned Case

Retrieved Case

New Case

Previous
Cases

Repaired case

Solved Case

Confirmed Solution

Proposed Solution

Fig. 5. Aamodt and Plaza's CBR cycle [13]

Retrieve : Case retrieval is a major phase in the CBR cycle where matching between
two cases plays a vital role. The retrieval step is essential especially in medical appli-
cations since lack of similar cases may lead to a less informed decision. The reliability
and accuracy of the diagnosis system depends on the storage of cases/experiences and
on the retrieval of all relevant cases and their ranking. The retrieved cases are ranked
on the basis of their similarity in matching the features of the new case and often the
highest ranked case is proposed as the solution to the new case. In new areas of medi-
cal research such as diagnosis of stress related to psychophysiological issues, the
domain knowledge is often not well understood. Therefore, retrieving a single match-
ing case as a proposed solution may not be sufficient for the decision support system
in this domain. The comparison of a new case with old cases from the case base could
be carried out applying different similarity matching algorithms. One of the com-
monly used similarity measurement techniques is the Nearest-neighbour algorithm
[12], [24]. A standard equation (shown in equation 1) for the nearest-neighbour is:

n

∑
=

Similarity ( C , S ) = w * sim ( C , S )

(1)

f f f

f 1

Where C is a current/target case, S is a stored case in the case base, w is the normal-
ized weight, n is the number of the attributes/features in each case, f is the index for
an individual attribute/feature and sim ( C f , S f ) is the local similarity function. Gener-
ally there are two ways to specify the values of weights for individual features. One
way is to define weights by experts in terms of the domain knowledge, while the other
is to learn or optimize weights using the case library as an information source. The
fuzzy similarity matching algorithm, which is another retrieval technique, is presented
in section 5.
Reuse and revise : The retrieved cases are sent to the reuse step (see Fig. 5) where
the solution of a past case can often be adapted to find a suitable solution for a new
case. A user can adapt solutions e.g. a combination of two solutions from the list of