0531426 20240103224308.920200805235959.9 Prague ÚTIA AV ČR, v.v.i 2020 10 s. P Research Report 2383 Deep Ensembles uncertainty neural networks cav_un_auth*0341005 Ulrych Lukáš UTIA-B Adaptivní systémy Department of Adaptive Systems AS AS Department of Adaptive Systems CZ Ústav teorie informace a automatizace AV ČR, v. v. i. cav_un_auth*0101207 Šmídl Václav UTIA-B Adaptivní systémy Department of Adaptive Systems AS AS Department of Adaptive Systems Ústav teorie informace a automatizace AV ČR, v. v. i. http://library.utia.cas.cz/separaty/2020/AS/smidl-0531426.pdf cav_un_auth*0374053 GA18-21409S GA ČR 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. IN 10000 10200 10201 2021 4 O 4o 20231122145040.2 RVO:67985556 http://hdl.handle.net/11104/0310095 S 2020 Soubory v repozitáři: 0531426.pdf 2020 Prague ÚTIA AV ČR, v.v.i