BigStat for Big Data: Big Data clustering through the BigStat SaaS platform

Christophe Biernacki (1, 2)

1 LPP - Laboratoire Paul Painlevé - UMR 8524
2 MODAL - MOdel for Data Analysis and Learning

BigStat is a web platform devoted to clustering of big data sets through two hosted software, MixtComp and BlockCluster. The former adresses mixed, missing and uncertain data in a moderate dimensional setting, whereas the latter is devoted to high dimensional data sets with non-mixed, non-missing and non-un certain data. Mathematical foundations of both rely on mixture models and related algorithms.

Mots clés: Model-based clustering; mixed and missing data; High dimension statistics; SaaS platform

Domaines: Méthodologie [stat.ME]

Christophe Biernacki. BigStat for Big Data: Big Data clustering through the BigStat SaaS platform. Journée scientifique « Big Data & Data science », Oct 2016, Tunis, Tunisia. ⟨hal-01420650⟩