Affinity Propagation Outlier Clustering

1:30 pm - 4:00 pm
Expo Hall/Career Fair

Objective: Inspiration
Audience Level: All
Session Type: Poster

In this paper I present an extension of affinity propagation which simultaneously clusters and detects outliers in data. The advantagesinclude: (i) the resulting clusters tend to be compact and semantically coherent (ii) the clusters are more robust against data perturbations and (iii) the outliers are contextualized by the clusters and are more interpretable.


Linsey Pang, Data Scientist/applied researcher, eBay