Combining Decision Trees and K-NN for Case-Based Planning
Sofia Benbelkacem, Baghdad Atmani, Mohamed Benamina · 2014
In everyday life, we are often faced with similar problems which we resolve with our experience.Case-based reasoning is a paradigm of problem solving based on past experience.Thus, case-based reasoning is considered as a valuable technique for the implementation of various tasks involving solving planning problem.Planning is considered as a decision support process designed to provide resources and required services to achieve specific objectives, allowing the selection of a better solution among several alternatives.However, we propose to exploit decision trees and k-NN combination to choose the most appropriate solutions.In a previous work [1], we have proposed a new planning approach guided by case-based reasoning and decision tree, called DTR, for case retrieval.In this paper, we use a classifier combination for similarity calculation in order to select the best solution to the target case.Thus, the use of the decision trees and k-NN combination allows improving the relevance of results and finding the most relevant cases.