286,462 Hits in 1.8 sec

A Random Matrix Theory Approach to Damping in Deep Learning [article]

Diego Granziol, Nicholas Baskerville
2022 arXiv   pre-print
We conjecture that the inherent difference in generalisation between adaptive and non-adaptive gradient methods in deep learning stems from the increased estimation noise in the flattest directions of  ...  effectively amplifying the noise-to-signal ratio and harming generalisation.  ...  with adaptive methods and closes the adaptive generalisation gap.  ... 
arXiv:2011.08181v5 fatcat:nyb2ckkmdndvfi4knz6ub6wqrm

Generalisable Cardiac Structure Segmentation via Attentional and Stacked Image Adaptation [article]

Hongwei Li, Jianguo Zhang, Bjoern Menze
2020 arXiv   pre-print
In this paper, we propose a generalisable segmentation framework for cardiac image segmentation in which multi-centre, multi-vendor, multi-disease datasets are involved.  ...  To mitigate these issues, domain adaptation methods have been proposed to generalise one algorithm trained on some datasets (source domain) with additional data (either labelled or not labelled) from another  ...  Data augmentation-based methods are further proposed to enhance the generalisability of cardiac image segmentation models [9, 2] .  ... 
arXiv:2008.01216v2 fatcat:csiaw3elcbdylhgpopdot5scxm

Microscope-AOtools: A generalised adaptive optics implementation [article]

Nicholas James Hall, Josh Titlow, Martin J Booth, Ian M Dobbie
2020 biorxiv/medrxiv   pre-print
Microscope-AOtools is a software package which allows for a simple, robust and generalised implementation of adaptive optics (AO) elements.  ...  These methods are constructed in a generalised manner that can utilise a range of adaptive optics elements, wavefront sensing techniques and sensorless AO correction methods.  ...  By constructing the calibration workflow in this generalised manner Microscope-AOtools can be used on any arbitrary Python Microscope compatible adaptive element with any wavefront sensing technique.  ... 
doi:10.1101/2020.06.18.158972 fatcat:7wuivtslozc7pl75il3oudm3ca

Generalised neuron-based adaptive power system stabiliser

D.K. Chaturvedi, O.P. Malik, P.K. Kalra
2004 IEE Proceedings - Generation Transmission and Distribution  
Taking benefit of these characteristics of the GN, a new generalised neuron-based adaptive power system stabiliser (GNPSS) is proposed.  ...  To overcome these drawbacks, a generalised neuron (GN) has been developed that requires much smaller training data and shorter training time.  ...  Conclusions Performance of an adaptive PSS based on a generalised neuron has been investigated in a single machine, infinite bus power system.  ... 
doi:10.1049/ip-gtd:20040084 fatcat:vlo66y5gwra4bptvtpi36y2ag4

Adaptive Control of Generalised Dynamically Substructured Systems

David P. Stoten, Jia-Ying Tu, Guang Li
2008 IFAC Proceedings Volumes  
This paper will present a generalised substructuring framework, using a practical example for illustration.  ...  Correspondingly, a linear substructuring control (LSC) strategy is presented, together with a development of the adaptive minimal control synthesis (MCS) algorithm.  ...  ACKNOWLEDGEMENT The authors gratefully acknowledge the support of the UK Engineering and Physical Science Research Council, grant number EP/D036917 (The adaptive control of generalised dynamically substructured  ... 
doi:10.3182/20080706-5-kr-1001.02385 fatcat:4mm3fsrntfby7nwbphtryfvcfy

Generalised array reconfiguration for adaptive beamforming by antenna selection

Xiangrong Wang, Elias Aboutanios, Moeness G. Amin
2015 2015 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)  
The performance of adaptive beamforming is not only dependent on the implemented algorithm, but also the array geometry [3] .  ...  CONCLUSION In this paper, we studied reconfigurable adaptive antenna arrays by antenna selection in multi-interference cases.  ... 
doi:10.1109/icassp.2015.7178417 dblp:conf/icassp/WangAA15 fatcat:w3z5p65dkfa4xa4wucvl2ywhmu

Generalised link-layer adaptation with costly power under higher-layer criteria

Virgilio Rodriguez
2011 34th IEEE Sarnoff Symposium  
It has long been recognised that a wireless communication system can be more efficient if link-layer parameters such as modulation order, symbol rate and packet size, are (adaptively) optimised.  ...  But a packet-oriented criterion for link adaptation seems more appropriate for practical communication networks fitted with strong error detection and a selective packet re-transmission mechanism.  ...  LINK ADAPTATION CRITERIA FOR DATA TRAFFIC WITH COSTLY POWER  ... 
doi:10.1109/sarnof.2011.5876443 fatcat:ocsmt2644fdvzdi4qunqcrhlta

A multivariate evolutionary generalised linear model framework with adaptive estimation for claims reserving [article]

Benjamin Avanzi, Gregory Clive Taylor, Phuong Anh Vu, Bernard Wong
2020 arXiv   pre-print
This is an adaptive estimation approach which updates evolving factors of the framework recursively over time.  ...  In this paper, we develop a multivariate evolutionary generalised linear model (GLM) framework for claims reserving, which allows for dynamic features of claims activity in conjunction with dependency  ...  Traditional reserving methods which use the assumption of similar claim activities across accident years, such as the traditional chain ladder algorithm and the traditional generalised linear model (GLM  ... 
arXiv:2004.06880v1 fatcat:zms2ydyvv5egfc2uvabjcegibq

Generalised Unsupervised Domain Adaptation of Neural Machine Translation with Cross-Lingual Data Selection [article]

Thuy-Trang Vu, Xuanli He, Dinh Phung, Gholamreza Haffari
2021 arXiv   pre-print
Our proposed method trains an adaptive layer on top of multilingual BERT by contrastive learning to align the representation between the source and target language.  ...  Once the in-domain data is detected by the classifier, the NMT model is then adapted to the new domain by jointly learning translation and domain discrimination tasks.  ...  Generalised Unsupervised Domain Adaptation Domain adaptation is an important problem in NMT as it is very expensive to obtain training data that are both large and relevant to all possible domains.  ... 
arXiv:2109.04292v1 fatcat:vhzoazxeujgo7mdmqh7igcyi7q

Adaptive optimisation of a generalised phase contrast beam shaping system

F. Kenny, F.S. Choi, J. Glückstad, M.J. Booth
2015 Optics Communications  
The generalised phase contrast (GPC) method provides versatile and efficient light shaping for a range of applications.  ...  We have implemented a generalised phase contrast system that used two passes on a single spatial light modulator (SLM).  ...  Conclusion In this paper we have presented a dynamic generalised phase contrast system using two passes on a single SLM.  ... 
doi:10.1016/j.optcom.2014.12.059 pmid:26089573 pmcid:PMC4459477 fatcat:37x57pjebnbhdcrm6ymqzqm7ra

Cross-speaker generalisation in two phoneme-level perceptual adaptation processes

Patrick van der Zande, Alexandra Jesse, Anne Cutler
2014 Journal of Phonetics  
Specifically, the experiments looked at whether visual information about the identity of the speaker could prevent generalisation from occurring.  ...  This study investigated whether these changes in the listeners' perceptual systems can generalise to the perception of speech from a novel speaker.  ...  Phonetic retuning generalises across speakers but not fully. On the other hand, for selective adaptation, generalisation occurs to its full extent.  ... 
doi:10.1016/j.wocn.2014.01.003 fatcat:4bhy4g7mgbajzminz6cguq7fyi

Sleep facilitates generalisation of accent adaptation to a new talker

Xin Xie, F. Sayako Earle, Emily B. Myers
2017 Language, Cognition and Neuroscience  
Lexically-guided phonetic retuning helps listeners adapt to the phonetic "fingerprint" of a talker.  ...  Experiment 1 showed adaptation for the trained talker and a weak transfer to the untrained talker.  ...  Immediate generalisation in talker adaptation Immediate generalisation is manifested in talker-specific adaptation.  ... 
doi:10.1080/23273798.2017.1369551 pmid:29372171 pmcid:PMC5778349 fatcat:x4o76b6ttbdedez6eh7npi73ay

Generalised Adaptive Harmony Search: A Comparative Analysis of Modern Harmony Search

Jaco Fourie, Richard Green, Zong Woo Geem
2013 Journal of Applied Mathematics  
Observations from this comparison led us to design a novel hybrid that combines the best attributes of these modern variants into a single optimiser called generalised adaptive harmony search.  ...  In this paper we compare the performance of three recent HS variants: exploratory harmony search, self-adaptive harmony search, and dynamic local-best harmony search.  ...  We suggested possible ways that some of these weaknesses can be addressed, and we proposed the generalised adaptive harmony search (GAHS) algorithm as a generalised version of AHS.  ... 
doi:10.1155/2013/380985 fatcat:wxd4mev6pfhgbkyc72yklodzby

Generalised parametric Rao test for multi-channel adaptive detection of range-spread targets

P. Wang, H. Li, T.R. Kavala, B. Himed
2012 IET Signal Processing  
To this end, a generalised parametric Rao (GP-Rao) test is developed by modelling the disturbance as a multi-channel auto-regressive process.  ...  The resulting detectors, including the generalised adaptive matched filter (GAMF) and the GLRT, use homogeneous training signals to estimate the unknown disturbance covariance matrix.  ...  The proposed Rao test for the range-spread target detection is referred to as the generalised parametric Rao (GP-Rao) test.  ... 
doi:10.1049/iet-spr.2011.0313 fatcat:xlbdbdktxzf2bjgz6y7k2bpnli

Generalisation and domain adaptation in GP with gradient descent for symbolic regression

Qi Chen, Bing Xue, Mengjie Zhang
2015 2015 IEEE Congress on Evolutionary Computation (CEC)  
well on the unseen data set in the same domain and cannot be adapted to unseen data in an extended domain.  ...  a series of experiment objectives to investigate the effectiveness and efficiency of GPGD with various settings for a set of symbolic regression problems applied to unseen data in the same domain and adapted  ...  Accordingly, applying gradient descent less often to few programs might enhance the generalisation and domain adaptation ability of the GPGD method.  ... 
doi:10.1109/cec.2015.7257017 dblp:conf/cec/ChenXZ15 fatcat:74dxvscezjaphaioayp552l6gu
« Previous Showing results 1 — 15 out of 286,462 results