search query: @indexterm data mining / total: 29
reference: 10 / 29
| Author: | Bohanec, M. Zupan, B. |
| Title: | A function-decomposition method for development of hierarchical multi-attribute decision models |
| Journal: | Decision Support Systems
2004 : JAN, VOL. 36:3, p. 215-233 |
| Index terms: | Decision making Data mining Methodology Models |
| Language: | eng |
| Abstract: | Function decomposition (hereafter as: f-d.) is a recent machine learning method developing a hierarchical structure from class-labeled data by discovering new aggregate attributes and their descriptions. It is shown that f-d. can be used to develop a hierarchical multi-attribute decision (here as: dec.) model (here as: dec-m./dec-ms.) from a given unstructured set of dec. examples. The method implemented in a system called HINT is experimentally evaluated on a real-world housing loans allocation problem and on the rediscovery of three hierarchical dec-ms. It is demonstrated that decomposition can discover meaningful and transparent dec-ms. of high classification accuracy. The effects of human interaction are specifically studied through either assistance or provision of background knowledge for f-d., and it is shown that this has a positive effect on both the comprehensibility and classification accuracy. |
SCIMA