SEMINAR 1 페이지

Various methods have been suggested for assigning attribute weights in case-based reasoning. However, previous methods of assigning attribute weights are limited when used to calculate the attribute weights of qualitative variables. Hence, this limits the types of variables that may be used in case-based reasoning.
To address this problem, this research proposes a new method, termed the dualoptimization method, that improves the calculation of both types of attribute weights and quantifies qualitative variables by allocating random variables and dual-optimizing attribute weights for random variables based on genetic algorithms.
Additionally, this research suggests the adaptation method for the GA-CBR cost model. By reflecting estimation error caused by differences between target and retrieved cases, a retrieved solution is adjusted to more appropriate.
To validate the proposed methods, we conducted two validations to estimate the construction cost of military barracks and public apartment projects. The validation results indicate that the dual-optimization methods are improved in terms of accuracy and stability compared to previous methods, and it can improve the accuracy of the model through the adaptation method.