Filters








32,886 Hits in 5.6 sec

Learning Additive Models Online with Fast Evaluating Kernels [chapter]

Mark Herbster
2001 Lecture Notes in Computer Science  
We develop three new techniques to build on the recent advances in online learning with kernels.  ...  First, we show that an exponential speed-up in prediction time per trial is possible for such algorithms as the Kernel-Adatron, the Kernel-Perceptron, and ROMMA for specific additive models.  ...  A portion of this research was undertaken while at the Computer Learning Research Centre at Royal Holloway University.  ... 
doi:10.1007/3-540-44581-1_29 fatcat:cgkijnnbtfc2fk4zpl75fzcdxq

Dual Memory Architectures for Fast Deep Learning of Stream Data via an Online-Incremental-Transfer Strategy [article]

Sang-Woo Lee, Min-Oh Heo, Jiwon Kim, Jeonghee Kim, Byoung-Tak Zhang
2015 arXiv   pre-print
We aim to train deep models from new data that consists of new classes, distributions, and tasks at minimal computational cost, which we call online deep learning.  ...  The proposed architecture consists of deep representation learners and fast learnable shallow kernel networks, both of which synergize to track the information of new data.  ...  Online Incremental Learning Algorithms for Fast Memory Shallow Kernel Networks on the Neural Networks We introduce the fast memory; shallow kernel networks on the neural networks.  ... 
arXiv:1506.04477v1 fatcat:ryij5slnsjhi7acauu3bhnfkry

A sliding-window online fast variational sparse Bayesian learning algorithm

Thomas Buchgraber, Dmitriy Shutin, H. Vincent Poor
2011 2011 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP)  
In this work a new online learning algorithm that uses automatic relevance determination (ARD) is proposed for fast adaptive nonlinear filtering.  ...  A sequential decision rule for inclusion or deletion of basis functions is obtained by applying a recently proposed fast variational sparse Bayesian learning (SBL) method.  ...  This method provides a fast decision rule for selecting model components and allows for addition of new basis functions as well as deletion of components currently in the model.  ... 
doi:10.1109/icassp.2011.5946747 dblp:conf/icassp/BuchgraberSP11 fatcat:3wzxop3vszejlnvdyrmostp3j4

KeLP: a Kernel-based Learning Platform for Natural Language Processing

Simone Filice, Giuseppe Castellucci, Danilo Croce, Roberto Basili
2015 Proceedings of ACL-IJCNLP 2015 System Demonstrations  
KeLP is a Java framework that enables fast and easy implementation of kernel functions over discrete data, such as strings, trees or graphs and their combination with standard vectorial kernels.  ...  Additionally, it provides several kernel-based algorithms, e.g., online and batch kernel machines for classification, regression and clustering, and a Java environment for easy implementation of new algorithms  ...  (ii) Additional-algorithm packages, e.g., online kernel machines, Nyström method (Williams and Seeger, 2001) and label sequence learning (Altun et al., 2003) , and (iii) additional-kernel packages, which  ... 
doi:10.3115/v1/p15-4004 dblp:conf/acl/FiliceCCB15 fatcat:7tk3wpimbjgunif4h6flggfwlu

Physics-informed Gaussian Process for Online Optimization of Particle Accelerators [article]

Adi Hanuka, X. Huang, J. Shtalenkova, D. Kennedy, A. Edelen, V. R. Lalchand, D. Ratner, J. Duris
2020 arXiv   pre-print
The ability to inform the machine-learning model with physics may have wide applications in science.  ...  Instead, here we use a fast approximate model from physics simulations to design the GP model.  ...  For systems with complex high-dimensional data structures, expressive kernels facilitate efficient learning from online acquired data.  ... 
arXiv:2009.03566v1 fatcat:tke6mmylqneijgonxz4qteryt4

Online Attentive Kernel-Based Temporal Difference Learning [article]

Guang Yang, Xingguo Chen, Shangdong Yang, Huihui Wang, Shaokang Dong, Yang Gao
2022 arXiv   pre-print
Experimental evaluations showed that OAKTD outperformed several Online Kernel-based Temporal Difference (OKTD) learning algorithms in addition to the Temporal Difference (TD) learning algorithm with Tile  ...  Therefore, a simpler and more adaptive approach is introduced to evaluate value function with the kernel-based model.  ...  To accurately evaluate performance of the attentive kernel-based model as sparse representation, we compared our proposed OAKTD 4 with OKTD, OSKTD and TD with Tile Coding.  ... 
arXiv:2201.09065v1 fatcat:yo66jiysnfheviqmileumklv4a

Dynamic Kernel Distillation for Efficient Pose Estimation in Videos [article]

Xuecheng Nie and Yuncheng Li and Linjie Luo and Ning Zhang and Jiashi Feng
2019 arXiv   pre-print
In particular, DKD introduces a light-weight distillator to online distill pose kernels via leveraging temporal cues from the previous frame in a one-shot feed-forward manner.  ...  To address this issue, we propose a novel Dynamic Kernel Distillation (DKD) model to facilitate small networks for estimating human poses in videos, thus significantly lifting the efficiency.  ...  It can also fast distill pose kernels in a one-shot manner, avoiding complex iterating utilized by previous online kernel learning models [4, 27] .  ... 
arXiv:1908.09216v1 fatcat:lqwphxayu5ddtgdb2mebip2gbu

Large-scale Online Kernel Learning with Random Feature Reparameterization

Tu Dinh Nguyen, Trung Le, Hung Bui, Dinh Phung
2017 Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence  
A typical online kernel learning method faces two fundamental issues: the complexity in dealing with a huge number of observed data points (a.k.a the curse of kernelization) and the difficulty in learning  ...  This view further inspires a direct application of stochastic gradient descent for updating our model under an online learning setting.  ...  This work was partially supported by the Australian Research Council (ARC) and the CoE in Machine Learning and Big Data.  ... 
doi:10.24963/ijcai.2017/354 dblp:conf/ijcai/NguyenLBP17 fatcat:ls6j4uo4ujajneh7fuym7udi24

Real-Time Local GP Model Learning [chapter]

Duy Nguyen-Tuong, Matthias Seeger, Jan Peters
2010 Studies in Computational Intelligence  
being sufficiently fast for real-time online learning.  ...  Due to reduced computational cost, local Gaussian processes (LGP) can be applied for larger sample-sizes and online learning.  ...  It also interesting to investigate further applications of the LGP in humanoid robotics with 35 of more DoFs and learning other types of the control such as operational space control.  ... 
doi:10.1007/978-3-642-05181-4_9 fatcat:liqxstsdxbftzbqupzcyoanw2q

Diversity-Promoting Deep Reinforcement Learning for Interactive Recommendation [article]

Yong Liu, Yinan Zhang, Qiong Wu, Chunyan Miao, Lizhen Cui, Binqiang Zhao, Yin Zhao, Lu Guan
2019 arXiv   pre-print
We performed extensive offline experiments as well as simulated online experiments with real world datasets to demonstrate the effectiveness of the proposed model.  ...  through an actor-critic reinforcement learning framework.  ...  For online evaluation, we build a simulator to compare D 2 RL with C 2 UCB.  ... 
arXiv:1903.07826v1 fatcat:s5nlfafmvjhmlct5gar2qkcrc4

Using model knowledge for learning inverse dynamics

Duy Nguyen-Tuong, Jan Peters
2010 2010 IEEE International Conference on Robotics and Automation  
The results show that the semiparametric models learned with rigid body dynamics as prior outperform the standard rigid body dynamics models on real data while generalizing better for unknown parts of  ...  We present two possible semiparametric regression approaches, where the knowledge of the physical model can either become part of the mean function or of the kernel in a nonparametric Gaussian process  ...  of the additional kernel k(·, ·) to the learning process.  ... 
doi:10.1109/robot.2010.5509858 dblp:conf/icra/Nguyen-TuongP10 fatcat:uftusyvf5vboxlrydvuhxehevu

Incremental online sparsification for model learning in real-time robot control

Duy Nguyen-Tuong, Jan Peters
2011 Neurocomputing  
In this paper, we propose a framework for online, incremental sparsification with a fixed budget designed for fast real-time model learning.  ...  In combination with an incremental learning approach such as incremental Gaussian process regression, we obtain a model approximation method which is applicable in real-time online learning.  ...  Acknowledgments We would like to thank Bernhard Schölkopf from Max Planck Institute for Biological Cybernetics for helpful discussions and for providing us with relevant machine learning literature of  ... 
doi:10.1016/j.neucom.2010.06.033 fatcat:2d5ofwppgfccnmsx5mnenezbtq

Robust and Real-Time Visual Tracking Based on Complementary Learners [chapter]

Xingzhou Luo, Dapeng Du, Gangshan Wu
2018 Lecture Notes in Computer Science  
In addition, we adopt the average peak-to-correlation energy to determine whether to activate and update an online CUR filter for re-detecting the target.  ...  In this paper, we exploit the low dimensional complementary features and an adaptive online detector with the average peak-to-correlation energy to improve tracking accuracy and time efficiency.  ...  Although several methods fuse multiple features or models to learn the target appearance model, the online models tend to drift due to fast motion and occlusion.  ... 
doi:10.1007/978-3-319-73600-6_19 fatcat:3jc54mtnsjdrjhxj4lw4oexss4

Long-term correlation tracking

Chao Ma, Xiaokang Yang, Chongyang Zhang, Ming-Hsuan Yang
2015 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR)  
In addition, we train an online random fern classifier to re-detect objects in case of tracking failure.  ...  We show that the correlation between temporal context considerably improves the accuracy and reliability for translation estimation, and it is effective to learn discriminative correlation filters from  ...  prevalent in online tracking with an online binary classifier.  ... 
doi:10.1109/cvpr.2015.7299177 dblp:conf/cvpr/MaYZY15 fatcat:45hg4wqx75esrnolt75um7yjs4

A Neural Network Approach for Online Nonlinear Neyman-Pearson Classification [article]

Basarbatu Can, Huseyin Ozkan
2020 arXiv   pre-print
As a result, we obtain an expedited online adaptation and powerful nonlinear Neyman-Pearson modeling.  ...  We sequentially learn the SLFN with stochastic gradient descent updates based on a Lagrangian NP objective.  ...  By enforcing a strict hard constraint on FPR with a very large κ ∞, one can immediately reject models (while evaluating various models) that violate FPR constraint with even a slight positive deviation  ... 
arXiv:2006.08001v2 fatcat:wk6hjb5uirfo5ecijzneib2ede
« Previous Showing results 1 — 15 out of 32,886 results