TY - GEN
T1 - Communication-efficient distributed online prediction by dynamic model synchronization
AU - Kamp, Michael
AU - Boley, Mario
AU - Keren, Daniel
AU - Schuster, Assaf
AU - Sharfman, Izchak
N1 - Funding Information: A preliminary extended abstract of this paper was presented at the BD3 workshop at VLDB’13. This research has been supported by the EU FP7-ICT-2013-11 under grant 619491 (FERARI).
PY - 2014
Y1 - 2014
N2 - We present the first protocol for distributed online prediction that aims to minimize online prediction loss and network communication at the same time. This protocol can be applied wherever a prediction-based service must be provided timely for each data point of a multitude of high frequency data streams, each of which is observed at a local node of some distributed system. Exemplary applications include social content recommendation and algorithmic trading. The challenge is to balance the joint predictive performance of the nodes by exchanging information between them, while not letting communication overhead deteriorate the responsiveness of the service. Technically, the proposed protocol is based on controlling the variance of the local models in a decentralized way. This approach retains the asymptotic optimal regret of previous algorithms. At the same time, it allows to substantially reduce network communication, and, in contrast to previous approaches, it remains applicable when the data is non-stationary and shows rapid concept drift. We demonstrate empirically that the protocol is able to hold up a high predictive performance using only a fraction of the communication required by benchmark methods.
AB - We present the first protocol for distributed online prediction that aims to minimize online prediction loss and network communication at the same time. This protocol can be applied wherever a prediction-based service must be provided timely for each data point of a multitude of high frequency data streams, each of which is observed at a local node of some distributed system. Exemplary applications include social content recommendation and algorithmic trading. The challenge is to balance the joint predictive performance of the nodes by exchanging information between them, while not letting communication overhead deteriorate the responsiveness of the service. Technically, the proposed protocol is based on controlling the variance of the local models in a decentralized way. This approach retains the asymptotic optimal regret of previous algorithms. At the same time, it allows to substantially reduce network communication, and, in contrast to previous approaches, it remains applicable when the data is non-stationary and shows rapid concept drift. We demonstrate empirically that the protocol is able to hold up a high predictive performance using only a fraction of the communication required by benchmark methods.
UR - http://www.scopus.com/inward/record.url?scp=84907009213&partnerID=8YFLogxK
U2 - https://doi.org/10.1007/978-3-662-44848-9_40
DO - https://doi.org/10.1007/978-3-662-44848-9_40
M3 - Conference contribution
SN - 9783662448472
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 623
EP - 639
BT - Machine Learning and Knowledge Discovery in Databases - European Conference, ECML PKDD 2014, Proceedings
PB - Springer Verlag
T2 - European Conference on Machine Learning and Knowledge Discovery in Databases, ECML PKDD 2014
Y2 - 15 September 2014 through 19 September 2014
ER -