Original title: DEnFi: Deep Ensemble Filter for Active Learning
Authors: Ulrych, Lukáš ; Šmídl, Václav
Document type: Research reports
Year: 2020
Language: eng
Series: Research Report, volume: 2383
Abstract: Deep Ensembles proved to be a one of the most accurate representation of uncertainty for deep neural networks. Their accuracy is beneficial in the task of active learning where unknown samples are selected for labeling based on the uncertainty of their prediction. Underestimation of the predictive uncertainty leads to poor exploration of the method. The main issue of deep ensembles is their computational cost since multiple complex networks have to be computed in parallel. In this paper, we propose to address this issue by taking advantage of the recursive nature of active learning. Specifically, we propose several methods how to generate initial values of an ensemble based of the previous ensemble. We provide comparison of the proposed strategies with existing methods on benchmark problems from Bayesian optimization and active classification. Practical benefits of the approach is demonstrated on example of learning ID of an IoT device from structured data using deep-set based networks.
Keywords: Deep Ensembles; neural networks; uncertainty
Project no.: GA18-21409S (CEP)
Funding provider: GA ČR

Institution: Institute of Information Theory and Automation AS ČR (web)
Document availability information: Fulltext is available at external website.
External URL: http://library.utia.cas.cz/separaty/2020/AS/smidl-0531426.pdf
Original record: http://hdl.handle.net/11104/0310095

Permalink: http://www.nusl.cz/ntk/nusl-438262


The record appears in these collections:
Research > Institutes ASCR > Institute of Information Theory and Automation
Reports > Research reports