|
Title:
|
AN AUTOMATIC METHOD TO ASSIGN LOCAL RISK |
|
Author(s):
|
J.l. Castro , M. Navarro , J.m. Sánchez , J.m. Zurita |
|
ISBN:
|
978-972-8924-60-7 |
|
Editors:
|
António Palma dos Reis |
|
Year:
|
2008 |
|
Edition:
|
Single |
|
Keywords:
|
Case-based reasoning, similarity, risk, recovery |
|
Type:
|
Full Paper |
|
First Page:
|
151 |
|
Last Page:
|
157 |
|
Language:
|
English |
|
Cover:
|
|
|
Full Contents:
|
click to dowload 
|
|
Paper Abstract:
|
Recovery is a very important stage in case-based reasoning (CBR) because part of the success of the system depends on
it. Many retrieval techniques have been developed in this area. These techniques use many different kinds of tools, from
nearest neighbour matching functions to neural networks or genetic algorithms. Most of these recovery techniques have
in common that they just take two factors into account before deciding what solution will be appropriate for solving the
problem: the similarity between attributes and the importance or weight that each attribute has in the problem. So, they
forget the suitability of the solutions for solving the problem. In order to help to retrieve not only the most similar case
but also the most appropriate solution, we defined the Local Risk variable whose values were assigned by a human
expert. To use this variable without the help of the expert, in this paper we present a new variable called Probabilistic
Local Risk, which will be assigned to each attribute automatically. To assign the value of this variable, we will use the
information provided by the case base. |
|
|
|
|
|
|
