Information Bottleneck Classification in Extremely Distributed Systems

• Published in: Entropy (Basel, Switzerland) Vol. 22; no. 11; p. 1237
• Main Authors: Ullmann, Denis, Rezaeifar, Shideh, Taran, Olga, Holotyak, Taras, Panos, Brandon, Voloshynovskiy, Slava
• Format: Journal Article
• Language: English
• Published: Switzerland MDPI 30.10.2020; MDPI AG
• Subjects: classification; decentralized model; deep networks; information bottleneck principle; rate-distortion theory; information bottleneck principle; decentralized model; deep networks; rate-distortion theory; classification

We present a new decentralized classification system based on a distributed architecture. This system consists of distributed nodes, each possessing their own datasets and computing modules, along with a centralized server, which provides probes to classification and aggregates the responses of node...