Statistically Significant Discriminative Patterns Searching

Hoang Son Pham (1) , Gwendal Virlet (2) , Dominique Lavenier (2) , Alexandre Termier (3)

1 ICTEAM - Institute of Information and Communication Technologies, Electronics and Applied Mathematics
2 GenScale - Scalable, Optimized and Parallel Algorithms for Genomics
3 LACODAM - Large Scale Collaborative Data Mining

In this paper, we propose a novel algorithm, named SSDPS, to discover patterns in two-class datasets. The SSDPS algorithm owes its eciency to an original enumeration strategy of the patterns, which allows to exploit some degrees of anti-monotonicity on the measures of discriminance and statistical significance. Experimental results demonstrate that the performance of the SSDPS algorithm is better than others. In addition, the number of generated patterns is much less than the number of the other algorithms. Experiment on real data also shows that SSDPS eciently detects multiple SNPs combinations in genetic data.

Mots clés: Anti-Monotonicity; Statistical Significance; Discriminative patterns; Discriminative Measures

Domaines: Base de données [cs.DB]; Intelligence artificielle [cs.AI]; Calcul parallèle, distribué et partagé [cs.DC]

Hoang Son Pham, Gwendal Virlet, Dominique Lavenier, Alexandre Termier. Statistically Significant Discriminative Patterns Searching. DaWaK 2019 - 21st International Conference on Big Data Analytics and Knowledge Discovery, Aug 2019, Linz, Austria. pp.105-115, ⟨10.1007/978-3-030-27520-4_8⟩. ⟨hal-02190793⟩