Timezone: »

 
Poster
Online Facility Location with Multiple Advice
Matteo Almanza · Flavio Chierichetti · Silvio Lattanzi · Alessandro Panconesi · Giuseppe Re

Fri Dec 10 08:30 AM -- 10:00 AM (PST) @

Clustering is a central topic in unsupervised learning and its online formulation has received a lot of attention in recent years. In this paper, we study the classic facility location problem in the presence of multiple machine-learned advice. We design an algorithm with provable performance guarantees such that, if the advice is good, it outperforms the best-known online algorithms for the problem, and if it is bad it still matches their performance.We complement our theoretical analysis with an in-depth study of the performance of our algorithm, showing its effectiveness on synthetic and real-world data sets.

Author Information

Matteo Almanza (Sapienza University of Rome)
Flavio Chierichetti (Sapienza University)
Silvio Lattanzi (Google Research)
Alessandro Panconesi (Sapienza, University of Rome)
Giuseppe Re (Sapienza University of Rome)

More from the Same Authors