Filters








403,957 Hits in 5.8 sec

Beyond Shared Hierarchies: Deep Multitask Learning through Soft Layer Ordering [article]

Elliot Meyerson, Risto Miikkulainen
2018 arXiv   pre-print
Such an organization significantly constricts the types of shared structure that can be learned.  ...  Existing deep multitask learning (MTL) approaches align layers shared between tasks in a parallel ordering.  ...  Generalizing the structure of shared layers. For clarity, in this paper all core layers in a given setup had the same shape.  ... 
arXiv:1711.00108v2 fatcat:zabncl3jfbe2dcrt636zna7ecm

Multi-Task Zipping via Layer-wise Neuron Sharing [article]

Xiaoxi He, Zimu Zhou, Lothar Thiele
2019 arXiv   pre-print
Central in MTZ is a layer-wise neuron sharing and incoming weight updating scheme that induces a minimal change in the error function.  ...  Evaluations show that MTZ is able to fully merge the hidden layers of two VGG-16 networks with a 3.18% increase in the test error averaged on ImageNet and CelebA, or share 39.61% parameters between the  ...  It selectively shares neurons and optimally updates their incoming weights on a layer basis to minimize the errors induced to each individual task.  ... 
arXiv:1805.09791v2 fatcat:mtsfbjfuqzawlixggzczyw22tq

Recurrent multiple shared layers in Depth for Neural Machine Translation [article]

GuoLiang Li, Yiyang Li
2021 arXiv   pre-print
To address the increasing of model parameters, we choose to share parameters in different recursive moments.  ...  Learning deeper models is usually a simple and effective approach to improve model performance, but deeper models have larger model parameters and are more difficult to train.  ...  To reduce the number of model parameters, parameter-sharing is an efficient and simple strategy.  ... 
arXiv:2108.10417v2 fatcat:kmv3f5xwwzf5jfshztmtsndvbq

Branched Multi-Task Networks: Deciding What Layers To Share [article]

Simon Vandenhende, Stamatios Georgoulis, Bert De Brabandere, Luc Van Gool
2020 arXiv   pre-print
Such ramified networks typically start with a number of shared layers, after which different tasks branch out into their own sequence of layers.  ...  Understandably, as the number of possible network configurations is combinatorially large, deciding what layers to share and where to branch out becomes cumbersome.  ...  We assume an appropriate structure for layer sharing to take the shape of a tree.  ... 
arXiv:1904.02920v5 fatcat:5hlzzhers5bonmbmrpwiw23k2y

Sharing e-Health Information through Ontological Layering

Pavandeep Kataria, Radmila Juric
2010 2010 43rd Hawaii International Conference on System Sciences  
Go-CID was introduced in [7] and its improved version was compared to an SA for retrievals of heterogonous medical sources in [6] .  ...  Sharing e-Health Information though Go-CID An instance of Go-CID In this section we apply the Scenario to illustrate the Go-CID layering as described in 1)-3) from section 3.  ... 
doi:10.1109/hicss.2010.338 dblp:conf/hicss/KatariaJ10 fatcat:af5tn7q2mngjrkbzkksvam3cou

Privacy-Preserving Image Sharing via Sparsifying Layers on Convolutional Groups [article]

Sohrab Ferdowsi, Behrooz Razeghi, Taras Holotyak, Flavio P. Calmon, Slava Voloshynovskiy
2020 arXiv   pre-print
We propose a practical framework to address the problem of privacy-aware image sharing in large-scale setups.  ...  The framework is tested on a large-scale database of images with public implementation available.  ...  The framework of Sparse Coding with Ambiguation (SCA) [10, 11] shares a compressed, but ambiguated representation of data within the public domain, while the users can benefit some utility given an authorized  ... 
arXiv:2002.01469v1 fatcat:kzi7dsyp35eaxb5iwf3g7dtvsm

Optimal Cross-Layer Design for Energy Efficient D2D Sharing Systems

Abdulrahman Alabbasi, Basem Shihada
2017 IEEE Transactions on Wireless Communications  
In this work, we propose a cross-layer design, which optimizes the energy efficiency of a potential future 5G spectrum-sharing environment, in two sharing scenarios.  ...  The cellular users, legacy users, are protected by an outage probability constraint.  ...  In the same context, several cross-layer design schemes (utilizing several layers ' parameters) have been proposed to improve the energy efficiency of different systems.  ... 
doi:10.1109/twc.2016.2631605 fatcat:gxg6q5j5wjhzvbmtjpluzglxsy

Privacy-Preserving Image Sharing Via Sparsifying Layers on Convolutional Groups

Sohrab Ferdowsi, Behrooz Razeghi, Taras Holotyak, Flavio P. Calmon, Slava Voloshynovskiy
2020 ICASSP 2020 - 2020 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)  
We propose a practical framework to address the problem of privacy-aware image sharing in large-scale setups.  ...  The framework is tested on a large-scale database of images with public implementation available.  ...  The framework of Sparse Coding with Ambiguation (SCA) [10, 11] shares a compressed, but ambiguated representation of data within the public domain, while the users can benefit some utility given an authorized  ... 
doi:10.1109/icassp40776.2020.9054046 dblp:conf/icassp/FerdowsiRHCV20 fatcat:4vdfpu7fsbalxn7klkoqkkxxra

Bayesian Optimization Based Efficient Layer Sharing for Incremental Learning

Bomi Kim, Taehyeon Kim, Yoonsik Choe
2021 Applied Sciences  
The layer sharing algorithm is one of the representative methods which leverages general knowledge by sharing some initial layers of the existing network.  ...  With the proposed method, the global optimal sharing layer can be found in only six or eight iterations without searching for all possible layer cases.  ...  MobileNetV2 is based on an inverted residual structure where the shortcut connections are between the thin bottleneck layer.  ... 
doi:10.3390/app11052171 doaj:6c24c7a9be324cf0b78a3d47fa729395 fatcat:kdqvvxxlofaa3krctlqti36luq

Privacy-Preserving Data Sharing using Multi-layer Access Control Model in Electronic Health Environment

Shekha Chenthara, Khandakar Ahmed, Frank Whittaker
2018 EAI Endorsed Transactions on Scalable Information Systems  
Electronic Health Data (EHD) is an emerging health information exchange model that facilitates healthcare providers and patients to efficiently store and share their private healthcare information from  ...  This paper proposes a multi-layer access control mechanism named MLAC Model to construct a secure and privacy-preserving EHR system that enables patients to share their data with stakeholders.  ...  Methodology and Conceptual Framework This work proposes a multi-layer access control mechanism, MLAC Model to construct a secure and privacy-preserving EHR system that enables patients to share their data  ... 
doi:10.4108/eai.13-7-2018.159356 fatcat:knojhxbexjh63mescdkupobk5a

Cross-language knowledge transfer using multilingual deep neural network with shared hidden layers

Jui-Ting Huang, Jinyu Li, Dong Yu, Li Deng, Yifan Gong
2013 2013 IEEE International Conference on Acoustics, Speech and Signal Processing  
Further, we show that the learned hidden layers sharing across languages can be transferred to improve recognition accuracy of new languages, with relative error reductions ranging from 6% to 28% against  ...  In this paper we propose a shared-hidden-layer multilingual DNN (SHL-MDNN), in which the hidden layers are made common across many languages while the softmax layers are made language dependent.  ...  Instead of enforcing the hard mapping, which could inaccurately represent the phonetic structure, the SHLs in the SHL-MDNN provides a natural structure to be shared across languages.  ... 
doi:10.1109/icassp.2013.6639081 dblp:conf/icassp/HuangLYDG13 fatcat:32gymepoubfzlku6mr4nsjxnlm

Research on the Prediction of A-Share "High Stock Dividend" Phenomenon—A Feature Adaptive Improved Multi-Layers Ensemble Model

Yi Fu, Bingwen Li, Jinshi Zhao, Qianwen Bi
2021 Entropy  
Using the A-share market data from 2010 to 2019 for simulation and evaluation, the proposed model improves the AUC (Area Under Curve) and F1 score by 0.173 and 0.303, respectively, compared to the baseline  ...  In this study, a new multi-layer stacking ensemble algorithm is proposed.  ...  China's A-share market is an important global investment market.  ... 
doi:10.3390/e23040416 pmid:33807274 fatcat:y7evb6v44rgt3iqfyv4ndjongm

Functional In-Place Update with Layered Datatype Sharing [chapter]

Michal Konečný
2003 Lecture Notes in Computer Science  
The system efficiently infers separation assertions as well as destruction and re-use effects for individual layers of recursive-type values.  ...  This research has been supported by the EPSRC grant GR/N28436/01 and by the EC Fifth Framework Programme project Mobile Resource Guarantees.  ...  Nevertheless, one can improve LFPL by relaxing its linearity in order to recognise more of the correct algorithms. This means losing the single pointer property and considering sharing.  ... 
doi:10.1007/3-540-44904-3_14 fatcat:bcbxvaikfbhyvf6c4rncvtstje

Secure sharing method of network data transmission based on multi-layer encryption technology

Jing Yanrong, Zhang Wenqian, Ge Lin, Li Nanfang, Shang Xiyuan, Ye Yun
2021 E3S Web of Conferences  
the network data transmission security sharing method based on multi-layer encryption technology are significantly improved compared with the traditional methods, and fully meet the research requirements  ...  based on multi-layer encryption technology is proposed.  ...  Realization of secure sharing of network data transmission The network data transmission security sharing framework consists of four levels: data layer, service layer, application layer and management  ... 
doi:10.1051/e3sconf/202125201027 doaj:be478c075739476493e711cefb82a29d fatcat:upl56a46nzc45d6nj3amgfgxbe

Exploration of multi-layered knowledge sharing participation: the roles of perceived benefits and costs

Mohammadbashir Sedighi, Sander van Splunter, Frances Brazier, Cees van Beers, Stephan Lukosch
2016 Journal of Knowledge Management  
The company has an electronic knowledge network between employees, in which they have the opportunity to share their knowledge in individual, group and organisation layers.  ...  Knowledge networks provide an opportunity for participants to interact and form relations in different layers of the knowledge exchange environment.  ... 
doi:10.1108/jkm-01-2016-0044 fatcat:lrordxoljfds5bvstmu4i22oqu
« Previous Showing results 1 — 15 out of 403,957 results