Tight Optimistic Estimates for Fast Subgroup Discovery

Großkreutz, Henrik; Rüping, Stefan; Wrobel, Stefan

doi:10.1007/978-3-540-87479-9_47

Cited by 65 publications

(84 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…An essential technique for efficient discovery showed to be pruning based on optimistic estimates [22,25]. As Grosskreutz et al showed, the efficiency of the pruning is strongly influenced by the tightness of the bounds [15]. A more general method to derive optimistic estimates for a whole class of interestingness measures, that is, convex measures, was introduced in [20] and later extended in [26].…”

Section: Related Workmentioning

confidence: 99%

“…A key technique to improve runtime performance of subgroup discovery in general is the application of optimistic estimates, that is, upper bounds for the interestingness of any specialization of the currently evaluated pattern. Although research has shown that improving the tightness of the utilized bounds improves the runtime performance substantially [15], there has been no extensive research so far concerning upper bounds for generalization aware interestingness measures beyond the trivial transfer of bounds for traditional measures.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Difference-Based Estimates for Generalization-Aware Subgroup Discovery

Lemmerich

Becker

Puppe

2013

Advanced Information Systems Engineering

View full text Add to dashboard Cite

Abstract. For the task of subgroup discovery, generalization-aware interesting measures that are based not only on the statistics of the patterns itself, but also on the statistics of their generalizations have recently been shown to be essential. A key technique to increase runtime performance of subgroup discovery algorithms is the application of optimistic estimates to limit the search space size. These are upper bounds for the interestingness that any specialization of the currently evaluated pattern may have. Until now these estimates are based on the anti-monotonicity of instances, which are covered by the current pattern. This neglects important properties of generalizations. Therefore, we present in this paper a new scheme of deriving optimistic estimates for generalization aware subgroup discovery, which is based on the instances by which patterns differ in comparison to their generalizations. We show, how this technique can be applied for the most popular interestingness measures for binary as well as for numeric target concepts. The novel bounds are incorporated in an efficient algorithm, which outperforms previous methods by up to an order of magnitude.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Difference-Based Estimates for Generalization-Aware Subgroup Discovery

Lemmerich

Becker

Puppe

2013

Advanced Information Systems Engineering

View full text Add to dashboard Cite

show abstract

“…All numerical attributes where discretized using minimal entropy discretization. As representative traditional subgroup miner we used the state-of-the-art algorithm Dpsubgroup [10]. All involved algorithms were implemented in Java and will be published on the author's webpage.…”

Section: Empirical Evaluationmentioning

confidence: 99%

“…Hence they create an anti-monotone search space that contains the family of all interesting descriptions. While our algorithms use equivalence classes instead of individual descriptions, all the optimistic estimator techniques including recent findings [10] can still be applied. Consequently, our algorithms always use a condensed version of traditional method's search spaces.…”

Section: Introductionmentioning

confidence: 99%

Non-redundant Subgroup Discovery Using a Closure System

Boley

Großkreutz

2009

Machine Learning and Knowledge Discovery in Databases

Self Cite

View full text Add to dashboard Cite

Abstract. Subgroup discovery is a local pattern discovery task, in which descriptions of subpopulations of a database are evaluated against some quality function. As standard quality functions are functions of the described subpopulation, we propose to search for equivalence classes of descriptions with respect to their extension in the database rather than individual descriptions. These equivalence classes have unique maximal representatives forming a closure system. We show that minimum cardinality representatives of each equivalence class can be found during the enumeration process of that closure system without additional cost, while finding a minimum representative of a single equivalence class is NP-hard. With several real-world datasets we demonstrate that search space and output are significantly reduced by considering equivalence classes instead of individual descriptions and that the minimum representatives constitute a family of subgroup descriptions that is of same or better expressive power than those generated by traditional methods.

show abstract

“…In order to obtain efficiency, these algorithms typically rely on top-down search combined with considerable pruning, exploiting either anti-monotonicity of the quality measure (e.g. frequency), or so-called optimistic estimates of the maximally attainable quality at every point in the search space (Grosskreutz et al 2008). With small datasets and simple tasks, these tricks work well and give complete solutions in reasonable time.…”

Section: Introductionmentioning

confidence: 99%

Diverse subgroup set discovery

Leeuwen

Knobbe

2012

Data Min Knowl Disc

View full text Add to dashboard Cite

Large data is challenging for most existing discovery algorithms, for several reasons. First of all, such data leads to enormous hypothesis spaces, making exhaustive search infeasible. Second, many variants of essentially the same pattern exist, due to (numeric) attributes of high cardinality, correlated attributes, and so on. This causes top-k mining algorithms to return highly redundant result sets, while ignoring many potentially interesting results. These problems are particularly apparent with subgroup discovery (SD) and its generalisation, exceptional model mining. To address this, we introduce subgroup set discovery: one should not consider individual subgroups, but sets of subgroups. We consider three degrees of redundancy, and propose corresponding heuristic selection strategies in order to eliminate redundancy. By incorporating these (generic) subgroup selection methods in a beam search, the aim is to improve the balance between exploration and exploitation. The proposed algorithm, dubbed DSSD for diverse subgroup set discovery, is experimentally evaluated and compared to existing approaches. For this, a variety of target types with corresponding datasets and quality measures is used. The subgroup sets that are discovered by the competing methods are evaluated primarily on the following three criteria: (1) diversity in the subgroup covers (exploration), (2) the maximum quality found (exploitation), and (3) runtime. The results show that DSSD outperforms each traditional SD method on all or a (non-empty) subset of these criteria, depending on the specific setting. The more complex the task, the larger the benefit of using our diverse heuristic search turns out to be.

show abstract

Tight Optimistic Estimates for Fast Subgroup Discovery

Cited by 65 publications

References 17 publications

Difference-Based Estimates for Generalization-Aware Subgroup Discovery

Difference-Based Estimates for Generalization-Aware Subgroup Discovery

Non-redundant Subgroup Discovery Using a Closure System

Diverse subgroup set discovery

Contact Info

Product

Resources

About