Brain-inspired Distributed Memorization Learning for Efficient Feature-free Unsupervised Domain Adaptation ...
Compared with gradient based artificial neural networks, biological neural networks usually show a more powerful generalization ability to quickly adapt to unknown environments without using any gradient back-propagation procedure. Inspired by the distributed memory mechanism of human brains, we pro...
Main Authors: | , , , , |
---|---|
Format: | Report |
Language: | unknown |
Published: |
arXiv
2024
|
Subjects: | |
Online Access: | https://dx.doi.org/10.48550/arxiv.2402.14598 https://arxiv.org/abs/2402.14598 |
id |
ftdatacite:10.48550/arxiv.2402.14598 |
---|---|
record_format |
openpolar |
spelling |
ftdatacite:10.48550/arxiv.2402.14598 2024-03-31T07:52:27+00:00 Brain-inspired Distributed Memorization Learning for Efficient Feature-free Unsupervised Domain Adaptation ... Lv, Jianming Liang, Depin Liang, Zequan Zhang, Yaobin Xia, Sijun 2024 https://dx.doi.org/10.48550/arxiv.2402.14598 https://arxiv.org/abs/2402.14598 unknown arXiv arXiv.org perpetual, non-exclusive license http://arxiv.org/licenses/nonexclusive-distrib/1.0/ Neural and Evolutionary Computing cs.NE Machine Learning cs.LG FOS Computer and information sciences article Preprint Article CreativeWork 2024 ftdatacite https://doi.org/10.48550/arxiv.2402.14598 2024-03-04T14:01:00Z Compared with gradient based artificial neural networks, biological neural networks usually show a more powerful generalization ability to quickly adapt to unknown environments without using any gradient back-propagation procedure. Inspired by the distributed memory mechanism of human brains, we propose a novel gradient-free Distributed Memorization Learning mechanism, namely DML, to support quick domain adaptation of transferred models. In particular, DML adopts randomly connected neurons to memorize the association of input signals, which are propagated as impulses, and makes the final decision by associating the distributed memories based on their confidence. More importantly, DML is able to perform reinforced memorization based on unlabeled data to quickly adapt to a new domain without heavy fine-tuning of deep features, which makes it very suitable for deploying on edge devices. Experiments based on four cross-domain real-world datasets show that DML can achieve superior performance of real-time domain ... : 15 pages,15 figures ... Report DML DataCite Metadata Store (German National Library of Science and Technology) |
institution |
Open Polar |
collection |
DataCite Metadata Store (German National Library of Science and Technology) |
op_collection_id |
ftdatacite |
language |
unknown |
topic |
Neural and Evolutionary Computing cs.NE Machine Learning cs.LG FOS Computer and information sciences |
spellingShingle |
Neural and Evolutionary Computing cs.NE Machine Learning cs.LG FOS Computer and information sciences Lv, Jianming Liang, Depin Liang, Zequan Zhang, Yaobin Xia, Sijun Brain-inspired Distributed Memorization Learning for Efficient Feature-free Unsupervised Domain Adaptation ... |
topic_facet |
Neural and Evolutionary Computing cs.NE Machine Learning cs.LG FOS Computer and information sciences |
description |
Compared with gradient based artificial neural networks, biological neural networks usually show a more powerful generalization ability to quickly adapt to unknown environments without using any gradient back-propagation procedure. Inspired by the distributed memory mechanism of human brains, we propose a novel gradient-free Distributed Memorization Learning mechanism, namely DML, to support quick domain adaptation of transferred models. In particular, DML adopts randomly connected neurons to memorize the association of input signals, which are propagated as impulses, and makes the final decision by associating the distributed memories based on their confidence. More importantly, DML is able to perform reinforced memorization based on unlabeled data to quickly adapt to a new domain without heavy fine-tuning of deep features, which makes it very suitable for deploying on edge devices. Experiments based on four cross-domain real-world datasets show that DML can achieve superior performance of real-time domain ... : 15 pages,15 figures ... |
format |
Report |
author |
Lv, Jianming Liang, Depin Liang, Zequan Zhang, Yaobin Xia, Sijun |
author_facet |
Lv, Jianming Liang, Depin Liang, Zequan Zhang, Yaobin Xia, Sijun |
author_sort |
Lv, Jianming |
title |
Brain-inspired Distributed Memorization Learning for Efficient Feature-free Unsupervised Domain Adaptation ... |
title_short |
Brain-inspired Distributed Memorization Learning for Efficient Feature-free Unsupervised Domain Adaptation ... |
title_full |
Brain-inspired Distributed Memorization Learning for Efficient Feature-free Unsupervised Domain Adaptation ... |
title_fullStr |
Brain-inspired Distributed Memorization Learning for Efficient Feature-free Unsupervised Domain Adaptation ... |
title_full_unstemmed |
Brain-inspired Distributed Memorization Learning for Efficient Feature-free Unsupervised Domain Adaptation ... |
title_sort |
brain-inspired distributed memorization learning for efficient feature-free unsupervised domain adaptation ... |
publisher |
arXiv |
publishDate |
2024 |
url |
https://dx.doi.org/10.48550/arxiv.2402.14598 https://arxiv.org/abs/2402.14598 |
genre |
DML |
genre_facet |
DML |
op_rights |
arXiv.org perpetual, non-exclusive license http://arxiv.org/licenses/nonexclusive-distrib/1.0/ |
op_doi |
https://doi.org/10.48550/arxiv.2402.14598 |
_version_ |
1795031582348673024 |