Abstract | ||
---|---|---|
We tackle the problem of finding association rules for quantitative data. Whereas most of the previous approaches operate on hyperrectangles, we propose a representation based on half-spaces. Consequently, the left-hand side and right-hand side of an association rule does not contain a conjunction of items or intervals, but a weighted sum of variables tested against a threshold. Since the downward closure property does not hold for such rules, we propose an optimization setting for finding locally optimal rules. A simple gradient descent algorithm optimizes a parameterized score function, where iterations optimizing the first separating hyperplane alternate with iterations optimizing the second. Experiments with two real-world data sets show that the approach finds non-random patterns and scales up well. We therefore propose quantitative association rules based on half-spaces as an interesting new class of patterns with a high potential for applications. |
Year | DOI | Venue |
---|---|---|
2004 | 10.1109/ICDM.2004.10038 | ICDM |
Keywords | Field | DocType |
data mining,gradient methods,optimisation,downward closure property,gradient descent algorithm,hyper rectangles,locally optimal rules,optimization,quantitative association rules | Downward closure,Data mining,Data set,Gradient descent,Mathematical optimization,Parameterized complexity,Computer science,Association rule learning,Artificial intelligence,Hyperplane,Score,Machine learning | Conference |
ISBN | Citations | PageRank |
0-7695-2142-8 | 19 | 1.37 |
References | Authors | |
10 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Ulrich Ruckert | 1 | 71 | 7.70 |
Lothar Richter | 2 | 98 | 6.32 |
Stefan Kramer | 3 | 1313 | 141.90 |