| Summary: | 碩士 === 國立中正大學 === 電機工程研究所 === 106 === Case-based reasoning (CBR) method, is based on an inspiration from experts of various domains, who prefer to rely on their experience from solving similar problems. If the attributes of cases are multivariate in time series, we must consider more than the distance between the attributes in the problem description part of two cases when measuring similarity. Moreover, to make sure that the retrieval solution can solve the target problem better, case adaptation is necessary. The goal of this study is to research case designing with multivariate attributes and implement case adaptation, with advantages of artificial neural networks and the concept of the case difference heuristic.
In data preprocessing, we emphasize the relevance between the cases with feature transformation and feature selection. By combining relevance with features to construct the case description of the target problem. The system retrieves the most similar case depending on attributes of the target case. The solution part of the retrieved case will then be refined after case adaptation.
The dataset used in our experiment comes from the UCI machine learning repository and is a collection of failure state of robot execution. Each set of data was obtained from a series of forces and torques measurements after robots facing the execution failures.
To show the effectiveness of our method, mean absolute error and Student’s t-test are used to determine that the CBR system with case adaptation is significantly better than the CBR system without case adaptation. As the result, this implies that case designed and case adaptation designed in this study is effective.
|
|---|
