Abstract | ||
---|---|---|
We present a novel data mining approach basedon decomposition. In order to analyze a givendataset, the method decomposes it to a hierarchyof smaller and less complex datasets that canbe analyzed independently. The method is experimentallyevaluated on a real-world housingloans allocation dataset, showing that the decompositioncan (1) discover meaningful intermediateconcepts, (2) decompose a relatively complexdataset to datasets that are easy to analyze andcomprehend, and (3) derive a... |
Year | Venue | Keywords |
---|---|---|
1997 | KDD | data mining |
Field | DocType | Citations |
Data mining,Computer science,Machine discovery,Human interaction,Artificial intelligence,Hierarchy,Classifier (linguistics),Machine learning | Conference | 5 |
PageRank | References | Authors |
0.57 | 5 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Blaz Zupan | 1 | 1277 | 102.37 |
Marko Bohanec | 2 | 334 | 48.69 |
Ivan Bratko | 3 | 1526 | 405.03 |
Bojan Cestnik | 4 | 716 | 262.57 |