| bibtype |
V -
Research Report
|
| ARLID |
0531426 |
| utime |
20240103224308.9 |
| mtime |
20200805235959.9 |
| title
(primary) (eng) |
DEnFi: Deep Ensemble Filter for Active Learning |
| publisher |
| place |
Prague |
| name |
ÚTIA AV ČR, v.v.i |
| pub_time |
2020 |
|
| specification |
| page_count |
10 s. |
| media_type |
P |
|
| edition |
| name |
Research Report |
| volume_id |
2383 |
|
| keyword |
Deep Ensembles |
| keyword |
uncertainty |
| keyword |
neural networks |
| author
(primary) |
| ARLID |
cav_un_auth*0341005 |
| name1 |
Ulrych |
| name2 |
Lukáš |
| institution |
UTIA-B |
| full_dept (cz) |
Adaptivní systémy |
| full_dept (eng) |
Department of Adaptive Systems |
| department (cz) |
AS |
| department (eng) |
AS |
| full_dept |
Department of Adaptive Systems |
| country |
CZ |
| fullinstit |
Ústav teorie informace a automatizace AV ČR, v. v. i. |
|
| author
|
| ARLID |
cav_un_auth*0101207 |
| name1 |
Šmídl |
| name2 |
Václav |
| institution |
UTIA-B |
| full_dept (cz) |
Adaptivní systémy |
| full_dept |
Department of Adaptive Systems |
| department (cz) |
AS |
| department |
AS |
| full_dept |
Department of Adaptive Systems |
| fullinstit |
Ústav teorie informace a automatizace AV ČR, v. v. i. |
|
| source |
|
| cas_special |
| project |
| ARLID |
cav_un_auth*0374053 |
| project_id |
GA18-21409S |
| agency |
GA ČR |
|
| abstract
(eng) |
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. |
| RIV |
IN |
| FORD0 |
10000 |
| FORD1 |
10200 |
| FORD2 |
10201 |
| reportyear |
2021 |
| mrcbC52 |
4 O 4o 20231122145040.2 |
| inst_support |
RVO:67985556 |
| permalink |
http://hdl.handle.net/11104/0310095 |
| confidential |
S |
| arlyear |
2020 |
| mrcbTft |
\nSoubory v repozitáři: 0531426.pdf |
| mrcbU10 |
2020 |
| mrcbU10 |
Prague ÚTIA AV ČR, v.v.i |
|